Discrete Choice Models
======================


.. _discrete_choice_example_notebook:

`Link to Notebook GitHub <https://github.com/statsmodels/statsmodels/blob/master/examples/notebooks/discrete_choice_example.ipynb>`_

.. raw:: html

   
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h2 id="Fair's-Affair-data">Fair's Affair data<a class="anchor-link" href="#Fair's-Affair-data">&#182;</a></h2>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>A survey of women only was conducted in 1974 by <em>Redbook</em> asking about extramarital affairs.</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[1]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="kn">from</span> <span class="nn">__future__</span> <span class="kn">import</span> <span class="n">print_function</span>
   <span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span>
   <span class="kn">from</span> <span class="nn">scipy</span> <span class="kn">import</span> <span class="n">stats</span>
   <span class="kn">import</span> <span class="nn">matplotlib.pyplot</span> <span class="kn">as</span> <span class="nn">plt</span>
   <span class="kn">import</span> <span class="nn">statsmodels.api</span> <span class="kn">as</span> <span class="nn">sm</span>
   <span class="kn">from</span> <span class="nn">statsmodels.formula.api</span> <span class="kn">import</span> <span class="n">logit</span><span class="p">,</span> <span class="n">probit</span><span class="p">,</span> <span class="n">poisson</span><span class="p">,</span> <span class="n">ols</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[2]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">fair</span><span class="o">.</span><span class="n">SOURCE</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   
   Fair, Ray. 1978. &quot;A Theory of Extramarital Affairs,&quot; &#96;Journal of Political
       Economy&#96;, February, 45-61.
   
   The data is available at http://fairmodel.econ.yale.edu/rayfair/pdf/2011b.htm
   
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[3]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span> <span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">fair</span><span class="o">.</span><span class="n">NOTE</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   ::
   
       Number of observations: 6366
       Number of variables: 9
       Variable name definitions:
   
           rate_marriage   : How rate marriage, 1 = very poor, 2 = poor, 3 = fair,
                           4 = good, 5 = very good
           age             : Age
           yrs_married     : No. years married. Interval approximations. See
                           original paper for detailed explanation.
           children        : No. children
           religious       : How relgious, 1 = not, 2 = mildly, 3 = fairly,
                           4 = strongly
           educ            : Level of education, 9 = grade school, 12 = high
                           school, 14 = some college, 16 = college graduate,
                           17 = some graduate school, 20 = advanced degree
           occupation      : 1 = student, 2 = farming, agriculture; semi-skilled,
                           or unskilled worker; 3 = white-colloar; 4 = teacher
                           counselor social worker, nurse; artist, writers;
                           technician, skilled worker, 5 = managerial,
                           administrative, business, 6 = professional with
                           advanced degree
           occupation_husb : Husband&apos;s occupation. Same as occupation.
           affairs         : measure of time spent in extramarital affairs
   
       See the original paper for more details.
   
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[4]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">dta</span> <span class="o">=</span> <span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">fair</span><span class="o">.</span><span class="n">load_pandas</span><span class="p">()</span><span class="o">.</span><span class="n">data</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[5]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">dta</span><span class="p">[</span><span class="s">&#39;affair&#39;</span><span class="p">]</span> <span class="o">=</span> <span class="p">(</span><span class="n">dta</span><span class="p">[</span><span class="s">&#39;affairs&#39;</span><span class="p">]</span> <span class="o">&gt;</span> <span class="mi">0</span><span class="p">)</span><span class="o">.</span><span class="n">astype</span><span class="p">(</span><span class="nb">float</span><span class="p">)</span>
   <span class="k">print</span><span class="p">(</span><span class="n">dta</span><span class="o">.</span><span class="n">head</span><span class="p">(</span><span class="mi">10</span><span class="p">))</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
      rate_marriage  age  yrs_married  children  religious  educ  occupation  \
   0              3   32          9.0       3.0          3    17           2   
   1              3   27         13.0       3.0          1    14           3   
   2              4   22          2.5       0.0          1    16           3   
   3              4   37         16.5       4.0          3    16           5   
   4              5   27          9.0       1.0          1    14           3   
   5              4   27          9.0       0.0          2    14           3   
   6              5   37         23.0       5.5          2    12           5   
   7              5   37         23.0       5.5          2    12           2   
   8              3   22          2.5       0.0          2    12           3   
   9              3   27          6.0       0.0          1    16           3   
   
      occupation_husb   affairs  affair  
   0                5  0.111111       1  
   1                4  3.230769       1  
   2                5  1.400000       1  
   3                5  0.727273       1  
   4                4  4.666666       1  
   5                4  4.666666       1  
   6                4  0.852174       1  
   7                3  1.826086       1  
   8                3  4.799999       1  
   9                5  1.333333       1  
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[6]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">dta</span><span class="o">.</span><span class="n">describe</span><span class="p">())</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
          rate_marriage          age  yrs_married     children    religious  \
   count    6366.000000  6366.000000  6366.000000  6366.000000  6366.000000   
   mean        4.109645    29.082862     9.009425     1.396874     2.426170   
   std         0.961430     6.847882     7.280120     1.433471     0.878369   
   min         1.000000    17.500000     0.500000     0.000000     1.000000   
   25%         4.000000    22.000000     2.500000     0.000000     2.000000   
   50%         4.000000    27.000000     6.000000     1.000000     2.000000   
   75%         5.000000    32.000000    16.500000     2.000000     3.000000   
   max         5.000000    42.000000    23.000000     5.500000     4.000000   
   
                 educ   occupation  occupation_husb      affairs       affair  
   count  6366.000000  6366.000000      6366.000000  6366.000000  6366.000000  
   mean     14.209865     3.424128         3.850141     0.705374     0.322495  
   std       2.178003     0.942399         1.346435     2.203374     0.467468  
   min       9.000000     1.000000         1.000000     0.000000     0.000000  
   25%      12.000000     3.000000         3.000000     0.000000     0.000000  
   50%      14.000000     3.000000         4.000000     0.000000     0.000000  
   75%      16.000000     4.000000         5.000000     0.484848     1.000000  
   max      20.000000     6.000000         6.000000    57.599991     1.000000  
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[7]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">affair_mod</span> <span class="o">=</span> <span class="n">logit</span><span class="p">(</span><span class="s">&quot;affair ~ occupation + educ + occupation_husb&quot;</span> 
                      <span class="s">&quot;+ rate_marriage + age + yrs_married + children&quot;</span>
                      <span class="s">&quot; + religious&quot;</span><span class="p">,</span> <span class="n">dta</span><span class="p">)</span><span class="o">.</span><span class="n">fit</span><span class="p">()</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   Optimization terminated successfully.
            Current function value: 0.545314
            Iterations 6
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[8]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">affair_mod</span><span class="o">.</span><span class="n">summary</span><span class="p">())</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
                              Logit Regression Results                           
   ==============================================================================
   Dep. Variable:                 affair   No. Observations:                 6366
   Model:                          Logit   Df Residuals:                     6357
   Method:                           MLE   Df Model:                            8
   Date:                Thu, 21 May 2015   Pseudo R-squ.:                  0.1327
   Time:                        05:57:15   Log-Likelihood:                -3471.5
   converged:                       True   LL-Null:                       -4002.5
                                           LLR p-value:                5.807e-224
   ===================================================================================
                         coef    std err          z      P&gt;|z|      [95.0% Conf. Int.]
   -----------------------------------------------------------------------------------
   Intercept           3.7257      0.299     12.470      0.000         3.140     4.311
   occupation          0.1602      0.034      4.717      0.000         0.094     0.227
   educ               -0.0392      0.015     -2.533      0.011        -0.070    -0.009
   occupation_husb     0.0124      0.023      0.541      0.589        -0.033     0.057
   rate_marriage      -0.7161      0.031    -22.784      0.000        -0.778    -0.655
   age                -0.0605      0.010     -5.885      0.000        -0.081    -0.040
   yrs_married         0.1100      0.011     10.054      0.000         0.089     0.131
   children           -0.0042      0.032     -0.134      0.893        -0.066     0.058
   religious          -0.3752      0.035    -10.792      0.000        -0.443    -0.307
   ===================================================================================
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>How well are we predicting?</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[9]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">affair_mod</span><span class="o">.</span><span class="n">pred_table</span><span class="p">()</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[9]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   array([[ 3882.,   431.],
          [ 1326.,   727.]])
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>The coefficients of the discrete choice model do not tell us much. What we&#39;re after is marginal effects.</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[10]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">mfx</span> <span class="o">=</span> <span class="n">affair_mod</span><span class="o">.</span><span class="n">get_margeff</span><span class="p">()</span>
   <span class="k">print</span><span class="p">(</span><span class="n">mfx</span><span class="o">.</span><span class="n">summary</span><span class="p">())</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
           Logit Marginal Effects       
   =====================================
   Dep. Variable:                 affair
   Method:                          dydx
   At:                           overall
   ===================================================================================
                        dy/dx    std err          z      P&gt;|z|      [95.0% Conf. Int.]
   -----------------------------------------------------------------------------------
   occupation          0.0293      0.006      4.744      0.000         0.017     0.041
   educ               -0.0072      0.003     -2.538      0.011        -0.013    -0.002
   occupation_husb     0.0023      0.004      0.541      0.589        -0.006     0.010
   rate_marriage      -0.1308      0.005    -26.891      0.000        -0.140    -0.121
   age                -0.0110      0.002     -5.937      0.000        -0.015    -0.007
   yrs_married         0.0201      0.002     10.327      0.000         0.016     0.024
   children           -0.0008      0.006     -0.134      0.893        -0.012     0.011
   religious          -0.0685      0.006    -11.119      0.000        -0.081    -0.056
   ===================================================================================
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[11]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">respondent1000</span> <span class="o">=</span> <span class="n">dta</span><span class="o">.</span><span class="n">ix</span><span class="p">[</span><span class="mi">1000</span><span class="p">]</span>
   <span class="k">print</span><span class="p">(</span><span class="n">respondent1000</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   rate_marriage       4.000000
   age                37.000000
   yrs_married        23.000000
   children            3.000000
   religious           3.000000
   educ               12.000000
   occupation          3.000000
   occupation_husb     4.000000
   affairs             0.521739
   affair              1.000000
   Name: 1000, dtype: float64
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[12]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">resp</span> <span class="o">=</span> <span class="nb">dict</span><span class="p">(</span><span class="nb">zip</span><span class="p">(</span><span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span><span class="mi">9</span><span class="p">),</span> <span class="n">respondent1000</span><span class="p">[[</span><span class="s">&quot;occupation&quot;</span><span class="p">,</span> <span class="s">&quot;educ&quot;</span><span class="p">,</span> 
                                               <span class="s">&quot;occupation_husb&quot;</span><span class="p">,</span> <span class="s">&quot;rate_marriage&quot;</span><span class="p">,</span> 
                                               <span class="s">&quot;age&quot;</span><span class="p">,</span> <span class="s">&quot;yrs_married&quot;</span><span class="p">,</span> <span class="s">&quot;children&quot;</span><span class="p">,</span> 
                                               <span class="s">&quot;religious&quot;</span><span class="p">]]</span><span class="o">.</span><span class="n">tolist</span><span class="p">()))</span>
   <span class="n">resp</span><span class="o">.</span><span class="n">update</span><span class="p">({</span><span class="mi">0</span> <span class="p">:</span> <span class="mi">1</span><span class="p">})</span>
   <span class="k">print</span><span class="p">(</span><span class="n">resp</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   {0: 1, 1: 3.0, 2: 12.0, 3: 4.0, 4: 4.0, 5: 37.0, 6: 23.0, 7: 3.0, 8: 3.0}
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[13]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">mfx</span> <span class="o">=</span> <span class="n">affair_mod</span><span class="o">.</span><span class="n">get_margeff</span><span class="p">(</span><span class="n">atexog</span><span class="o">=</span><span class="n">resp</span><span class="p">)</span>
   <span class="k">print</span><span class="p">(</span><span class="n">mfx</span><span class="o">.</span><span class="n">summary</span><span class="p">())</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
           Logit Marginal Effects       
   =====================================
   Dep. Variable:                 affair
   Method:                          dydx
   At:                           overall
   ===================================================================================
                        dy/dx    std err          z      P&gt;|z|      [95.0% Conf. Int.]
   -----------------------------------------------------------------------------------
   occupation          0.0400      0.008      4.711      0.000         0.023     0.057
   educ               -0.0098      0.004     -2.537      0.011        -0.017    -0.002
   occupation_husb     0.0031      0.006      0.541      0.589        -0.008     0.014
   rate_marriage      -0.1788      0.008    -22.743      0.000        -0.194    -0.163
   age                -0.0151      0.003     -5.928      0.000        -0.020    -0.010
   yrs_married         0.0275      0.003     10.256      0.000         0.022     0.033
   children           -0.0011      0.008     -0.134      0.893        -0.017     0.014
   religious          -0.0937      0.009    -10.722      0.000        -0.111    -0.077
   ===================================================================================
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[14]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">affair_mod</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span><span class="n">respondent1000</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[14]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   array([ 0.51878156])
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[15]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">affair_mod</span><span class="o">.</span><span class="n">fittedvalues</span><span class="p">[</span><span class="mi">1000</span><span class="p">]</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[15]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0.075161592850599757
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[16]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">affair_mod</span><span class="o">.</span><span class="n">model</span><span class="o">.</span><span class="n">cdf</span><span class="p">(</span><span class="n">affair_mod</span><span class="o">.</span><span class="n">fittedvalues</span><span class="p">[</span><span class="mi">1000</span><span class="p">])</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[16]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0.51878155721215624
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>The &quot;correct&quot; model here is likely the Tobit model. We have an work in progress branch &quot;tobit-model&quot; on github, if anyone is interested in censored regression models.</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h3 id="Exercise:-Logit-vs-Probit">Exercise: Logit vs Probit<a class="anchor-link" href="#Exercise:-Logit-vs-Probit">&#182;</a></h3>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[17]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">)</span>
   <span class="n">support</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">linspace</span><span class="p">(</span><span class="o">-</span><span class="mi">6</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">1000</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">support</span><span class="p">,</span> <span class="n">stats</span><span class="o">.</span><span class="n">logistic</span><span class="o">.</span><span class="n">cdf</span><span class="p">(</span><span class="n">support</span><span class="p">),</span> <span class="s">&#39;r-&#39;</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="s">&#39;Logistic&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">support</span><span class="p">,</span> <span class="n">stats</span><span class="o">.</span><span class="n">norm</span><span class="o">.</span><span class="n">cdf</span><span class="p">(</span><span class="n">support</span><span class="p">),</span> <span class="n">label</span><span class="o">=</span><span class="s">&#39;Probit&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">legend</span><span class="p">();</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAsEAAAGjCAYAAADNZci4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xm8VfP6wPHPKhWSNBANZAiZM5RrPOaM4eqaLjJPIS66
   xXXjGq7ZjZtCpNRNxjJPyZwmIgpNGiS/ijRPZ/3++DacUmfo7L3X3md/3q/Xeu3hrLPWczrKc77n
   +T5PFMcxkiRJUj6plHQAkiRJUqaZBEuSJCnvmARLkiQp75gES5IkKe+YBEuSJCnvmARLkiQp75SY
   BEdR9GQURdOjKPq6mHM6R1H0QxRFI6MoapbaECVJkqTUKs1K8FNAy3V9MIqi44Ad4jhuAlwCPJqi
   2CRJkqS0KDEJjuP4I+DXYk45CXh6+bmfA5tFUVQvNeFJkiRJqZeKmuAGwOQir6cADVNwXUmSJCkt
   UrUxLlrjtbOYJUmSlLU2SME1pgKNirxuuPy91URRZGIsSZKkjIjjeM1F2tWkIgkeALQF+kZRtD/w
   WxzH09cRTApupyR06tSJTp06JR2G1oPfu9zm9y93Zep7t2wZ/PQTTJgQjilTYNo0+Pnn8Lji2GAD
   qFMHateGWrXCY9Hnm20GNWpA9eqwySbhcc3nG20EUbFpRcVRof/uFRbCvHmrjrlzi3+9YAEsXBiO
   os/X9V7R10uWQLVqsOGG4ahWDapWDUeVKsU/L+17K56vODbYgOj000v8YygxCY6i6H/AoUDdKIom
   A/8EqgDEcdwtjuPXoyg6LoqiscA84PxyfWMkSdJq4hh++QW++Qa+/TYcY8eGpHfy5JDEbrcdbLst
   NGwIO+4Ihx4KW20FW24ZHqtXT/qrULnEcUguZ88Ox++/r3q+ruP332HOnD8mtgsXwsYbr/rpZs2f
   dtb2um7dVYnsimOjjUp+XbVqMj85pSIJjuP4zFKc07aUIUmSpGLEcUhwhw4Nx4gRIfmNY9h1V9hl
   F2jaFE48MSS922wTcg3lkPnzYebM0h2zZoVj9myoXBlq1lx1bLrp6q9r1oStt1799Yrl/aKJ7UYb
   QSXnpaWiHEJ5oKCgIOkQtJ783uU2v3+5q7Tfu4UL4fPP4f334ZNPYNiwkNvst184brkFdtsNttgi
   f0oRskGZ/u7FcVhpnT491KEUfVzzvV9+CeUIdeqs/WjYEPbcMyzvr3ivdu2Q0FarlravNx9FmarT
   jaIotiZYkpTv4hi++AJefx0GDoQhQ8Lq7mGHwcEHh8S3nt32s0ccw4wZodh68uTwWPT55Mmh6BrC
   N27LLYt/3GKLsBrrTzRpFUVRiRvjTIIlSUqzhQvDSu+AAfDKK6Ec84QT4IgjQuK76aZJR5jHVhRc
   jx+/+jFxYkhwp04N37CGDaFRo/C45vP69UPZwTpEJrxptbb80iRYkqSELFsWVnp794b+/UNJw0kn
   hWOnnZKOLs8UFsKkSTBmDHz//erJ7oQJIcndbrvVj8aNVyW6G29crtsvT8hS87VoNev6szUJliQp
   w777Dh57DPr0CfnTWWfBGWeEDg1Ks/nz4YcfYPTokPCuOL7/PtTV7rxzaJ2x/farkt1tt037UrxJ
   cPqYBEuSlKAlS0Kpw6OPwtdfwwUXQJs2rvimTWFhWMH96isYOTIcX30VGiZvv31IdoseO+1UbLlC
   upkEp49JsCRJCZg7Fx5/HB54ICwoXn45nHqqm/hTatGikOCOGLEq4f3669AtYc89w7HHHuFxhx3C
   VJAsYxKcPuVJgrPvvxRJkrLczJnQuTN06RK6OvTvD3vvnXRUFUBhYShnGDJk1fH11yG53Wcf2Gsv
   aN06JL116iQdrdZw+eWX06BBA26++eYyfd6kSZPYdddd+f333zO6idCVYEmSSmnOnLDq27kznHIK
   3HhjKDHVepo7Fz77DD76CD79NDRJrlULmjdfdTRrFoY85LBsXQlu3Lgx3bt354gjjsj4fZ988kkO
   P/zwcl/LlWBJktJo8eJQ9nD77aGt2dChYU+VymjGDPj445D0fvRRmP/crFnoE3fttSHp3XzzpKPM
   G1EUJdK+LVt+KHBmniRJxXjzzTCu+JVXwoCLZ54xAS61efPCH9q114YecdtvD127hk4N998fkuKP
   PoI774TjjzcBzgKLFi2iXbt2NGjQgAYNGnDttdeyePHilR+/5557qF+/Pg0bNuSJJ56gUqVKjB8/
   HoA2bdrwj3/8A4AZM2ZwwgknUKtWLerUqcMhhxxCHMecc845TJo0iRNPPJEaNWpw3333MXHiRCpV
   qkRhYSEAs2bN4vzzz6dBgwbUrl2bU045JS1fqyvBkiStxZQpIXcbMQIeeQSOPTbpiHLAsmUwfDi8
   8044hg8PtbxHHQVPPhkKp7Nw45pWueOOOxgyZAgjR44EoFWrVtx+++3cdtttvPnmmzz44IMMHDiQ
   xo0bc/HFF6/2uUVXlu+//34aNWrEjBkzABg8eDBRFNGrVy8+/vhjunfvvrIcYuLEiatd55xzzmHT
   TTfl22+/pXr16nz22Wdp+VpdCZYkqYhly+DBB8MerF12gVGjTICLNXs2PPssnH12GAl8wQVhhffG
   G8M44UGD4KabQqmDCfDaRVFqjhTo06cPt9xyC3Xr1qVu3br885//pFevXgD069ePCy64gKZNm7LR
   Rhtx6623rvM6VatWZdq0aUycOJHKlStz4IEHlur+06ZN480336Rr167UrFmTDTbYgIMPPjglX9ua
   /K9RkqTlxo2D884Ludqnn7rpbZ0mTAj1IQMGhA4OBx8cRuHdfXeYEKKyyYL62BV++uknttlmm5Wv
   t956a3766ScgJKjNmzdf+bGGa/ler6j1veGGG+jUqRNHH300AJdccgnt27cv8f6TJ0+mdu3a1KxZ
   s1xfR2m4EixJyntxDN26wf77w5//HMYdmwCv4fvvw87APfeEFi3gyy+hbduw2vvaa3DppSbAFUD9
   +vVXK0+YNGkSDRo0AGCrrbZi8uTJKz9W9PmaNtlkE+677z7GjRvHgAEDeOCBB3j//fcBit2M16hR
   I2bNmsXs2bPL+ZWUzCRYkpTXfv01DLh47DH44INQB1zJ/zsG48bBXXeFDg6HHgrTp4cC6WnTQo3v
   ySdD9epJR6lyWLx4MQsXLlx5nHnmmdx+++3MmDGDGTNmcNttt/HXv/4VgL/85S889dRTjBkzhvnz
   5/Ovf/1rtWsV7fjw6quvMnbsWOI4ZtNNN6Vy5cpUWv4Xq169eowbN26t8Wy11VYce+yxXHHFFfz2
   228sWbKEDz/8MC1fu3/NJUl5a9iwsG+rUaNQ/rDLLklHlAV+/jl0bth3XzjgAJg8ORRJT5kCDz8c
   Sh8qV046SqXIcccdx8Ybb7zyWLRoEfvuuy977LEHe+yxB/vuu+/K4RctW7bk6quv5rDDDmPHHXfk
   T3/6EwDVlo9ILLoxbuzYsRx11FHUqFGDAw44gCuvvJJDDz0UgA4dOnD77bdTq1YtHnjggZWfu0Kv
   Xr2oUqUKO++8M/Xq1aNz585p+dodliFJyktdu8Itt4Spb6edlnQ0CVu0CF59FXr0CC3LTj4Z/vpX
   KChwM1sKZEtf3FQbPXo0u+++O4sXL165yptpDsuQJKmUliyBa64JTQs+/TRM5M1bX3wRyhr69g19
   fNu0gf/9L+cntCl9XnrpJY477jjmz59P+/btOemkkxJLgMsrN6OWJGk9zJoV2p1NnBim9eZlArxg
   ATz9dNgFePLJULduGIH3/vuhNYYJsIrx2GOPUa9ePXbYYQeqVKnCo48+mnRI681yCElSXhg/Hlq2
   hBNPhHvuycOy1rFjQw3I00+Het8rroDjjsvDP4jMq6jlENmgPOUQrgRLkiq8kSPDfq527cKer7zJ
   ++IY3n03LH8fcED4wj//HN54I/w0kDd/ENIfWRMsSarQPvgAWreG//43POaFJUvCFLf77gvP//Y3
   eOkl2HDDpCOTsoZJsCSpwnr11TDFt29fOPzwpKPJgNmzQ8Pjzp3DtI+77oJjjrHxsbQWJsGSpApp
   wAC4+OIwzGy//ZKOJs1+/RUeeigsdx9zTPjimzVLOiopq/mjoSSpwunfP08S4Jkz4eabQ5uLKVNC
   vW/v3ibAUimYBEuSKpSXX4ZLLw17v/bdN+lo0mTGDOjYMZQ8TJ8eRt917w7bb590ZMpDlSpVYvz4
   8Wv9WO/evTnmmGMyHFHpmARLkiqMd9+FSy6B11+HvfdOOpo0mDcPbr8ddt45rAIPHw6PPw7bbpt0
   ZMpBjRs3ZuONN6ZGjRpsueWWnH/++cybNy+l9zj77LN56623Vr4uLmHONJNgSVKFMGQInHUWvPBC
   BUyAlyyBbt2gSRMYNQoGDw6vGzdOOjLlsCiKePXVV5kzZw4jRoxg2LBh3H777auds3Tp0pTfN1t6
   JpsES5Jy3rffwkknhQnABx+cdDQpFMchq99tN+jXL2x469s3T0fdKZ3q16/Psccey6hRo6hUqRJd
   unShSZMm7LTTTgA8/vjjNGnShDp16tCqVSumTZu22ue/9tprbL/99my++ebceOONKxPdHj16cPDy
   v5SHHHIIAHvuuSc1atTgueeey+BX+EcmwZKknDZlSpgEd999cMIJSUeTQiNHQkEB3HYbPPxwqPWo
   sEXOSsqKZHXy5Mm8/vrrNFu+qbJ///4MHTqUb7/9loEDB9KxY0eee+45pk2bxjbbbMMZZ5yx2nVe
   fvllhg8fzogRI+jfvz9PPvnkH+714YcfAvDVV18xZ84cWifcuNskWJKUs+bODYPPrrwS/vrXpKNJ
   kZkzw0jjo4+GM8+EESPC86jYCbDKYVGUmqOs4jjm5JNPplatWhx88MEUFBTQsWNHADp06MBmm21G
   tWrV6N27NxdeeCF77bUXVatW5a677uKzzz5j0qRJK6/Vvn17NttsMxo1akS7du343//+l6o/nrSx
   T7AkKSctWxZyxH32gRtvTDqaFFi6NNT53nornH46jB4NtWsnHZUyIKkS2SiK6N+/P4evZZJMo0aN
   Vj6fNm0a+xb5LUT16tWpU6cOU6dOZeutt/7D+VtvvTU//fRTGiNPDZNgSVJOuu46WLAAHn20AiyS
   Dh8eGhvXrAnvvQe77550RMpzUZG/VPXr12fixIkrX8+bN4+ZM2fSoEGDle9NmjSJpk2brnxe9GPZ
   ynIISVLO6doV3n4bnn8eqlRJOppymDs3ZPPHHw/t2sHAgSbAyjpnnnkmTz31FCNHjmTRokV07NiR
   /ffff+UqMMB9993Hb7/9xuTJk+ncuTOnn376Wq9Vr149xo0bl6nQi2USLEnKKZ99BrfcEholbLZZ
   0tGUw2uvha4PM2eGtmfnnlsBlrRVEURr/Hd4xBFH8K9//Ys///nP1K9fnwkTJtC3b9/VzmnVqhX7
   7LMPzZo144QTTuDCCy9cea2i1+vUqRPnnXcetWrV4vnnn0//F1OMKFO92qIoirOlL5wkKTf9/HMY
   g/zoozncCWLGDGjbNkx569oVjjwy6YiUZlEUZU1v3IpmXX+2y98v9qdKV4IlSTlhyRJo3RouuiiH
   E+ABA2CPPaBBA/j6axNgKUGuBEuSckK7djBuHPTvD5VybQln9uzwBXz4IfToUcEmeqgkrgSnjyvB
   kqQKrX//cPTsmYMJ8Lvvhs1uG24YBmCYAEtZwRZpkqSsNnkyXHJJSIJr1Uo6mjJYvBg6dAjjjp94
   Ao45JumIJBVhEixJylpLl4aBGNddB/vvn3Q0ZTB2LJxxRqj9/fJLqFMn6YgkrSHXfqkkScojt94K
   1avDDTckHUkZ9O4Nf/oTtGkDL79sAixlKVeCJUlZ6aOPoHt3+OKLHKkDnjsXrroqNDJ+5x3Ya6+k
   I1IWWbP3rpJnEixJyjpz54aF1K5doV69pKMphe+/h1NOCU2Mhw2DTTZJOiJlETtDZKdc+NlakpRn
   brgBDjkETjop6UhKoX9/OOgguOYaeOopE2ApR7gSLEnKKm+9Ba+/Dl99lXQkJVi2DDp1gqefhlde
   gRYtko5IUhmYBEuSssavv4aJcE89BTVrJh1NMWbNgrPOgoULYejQHKnZkFSU5RCSpKxxzTXQqlWW
   TxP+9ttQ+7vrrmEQhgmwlJNcCZYkZYXXXoNPPsnyMoi33oJzzoH77oNzz006GknlYBIsSUrc3Llw
   5ZVhsFr16klHsw5dusBtt8ELLzj6WKoAoky17YiiKLZFiCRpbf72N/jlF+jVK+lI1mLp0jCy7p13
   4NVXYfvtk45IUgmiKCKO42KbM7sSLElK1IgR8MwzMGpU0pGsxe+/h/HHS5eGIRibbZZ0RJJSxI1x
   kqTELF0Kl1wCd98Nm2+edDRrmD4dCgqgUaPQs80EWKpQTIIlSYl5+GHYdFM477ykI1nD2LFwwAFw
   8slhbN0G/uJUqmisCZYkJWLyZGjWDD79FHbcMeloihgxAk44Af75T7j00qSjkbQeSlMTbBIsSUrE
   mWdCkyah4ULWePfdMASjWzc45ZSko5G0nkyCJUlZ6cMP4a9/hTFjYOONk45muX79oG1beP55OOSQ
   pKORVA52h5AkZZ1ly+Dqq8O8iaxJgHv0gI4dw0rwHnskHY2kDDAJliRl1OOPh0YLrVsnHclyXbvC
   HXfAwIGw885JRyMpQyyHkCRlzKxZ0LQpvP027Lln0tEADz0E//lPWAF2CIZUYVgTLEnKKm3bQhzD
   f/+bdCTAnXfCU0/Be+/B1lsnHY2kFLImWJKUNUaNCnvPRo9OOJA4Du3Pnn8ePvgA6tdPOCBJSTAJ
   liRlxN//Hvae1amTcCC33govvQSDBsEWWyQcjKSkmARLktLu/ffDCvCLLyYcyJ13huVoE2Ap75kE
   S5LSqrAQbrwx5J9VqyYYyH33hVZoH3xgAizJJFiSlF79+kEUJdwSrXNnePTRkABvtVWCgUjKFnaH
   kCSlzaJFoSXak09CQUFCQXTrBnfdFRLgbbZJKAhJmVSa7hCVSnGRllEUjYmi6Icoitqv5eN1oyh6
   M4qiL6MoGhVFUZtyxCxJqkC6dIFdd00wAe7ZE26/PbRBMwGWVESxK8FRFFUGvgOOBKYCQ4Ez4zge
   XeScTkC1OI47RFFUd/n59eI4XrrGtVwJlqQ88ttvsOOOYVPcrrsmEMArr8DFF4cAmjZNIABJSUnF
   SnBzYGwcxxPjOF4C9AVarXHONGDT5c83BWaumQBLkvLPAw/A8ccnlAB//DFccAEMGGACLGmtStoY
   1wCYXOT1FKDFGuc8DgyMougnoAbwl9SFJ0nKRTNmhKlww4YlcPOvvoI//xn69IHmzRMIQFIuKGkl
   uDT1Cx2BL+M4rg/sBfw3iqIa5Y5MkpSz7r0X/vIX2HbbDN94/Hg49lh4+GE46qgM31xSLilpJXgq
   0KjI60aE1eCiDgDuAIjjeFwURROAnYA//PzfqVOnlc8LCgooSGynhCQpXX7+GZ54AkaOTODGRx8N
   N90UMnBJeWPQoEEMGjSoTJ9T0sa4DQgb3Y4AfgKG8MeNcQ8As+M4vjWKonrAcGCPOI5nrXEtN8ZJ
   Uh5o1y48PvRQBm86dy4ceiiceCIUWXCRlJ9KszGuxD7BURQdCzwEVAa6x3F8VxRFlwLEcdxteUeI
   p4CtCeUVd8Vx3Gct1zEJlqQKbsoU2HNP+OYb2HLLDN106VI4+eRww8cfD5M5JOW1lCTBKQzGJFiS
   KrjLL4dNN4W7787QDeMY2raFH36A116DKlUydGNJ2aw0SbBjkyVJKfHjj2FE8nffZfCmDz4IH30U
   DhNgSWVgEixJSom774ZLL4W6dTN0wxdeCM2IP/sMatbM0E0lVRSWQ0iSym3qVNh997AKvPnmGbjh
   4MFw0knw1lvQrFkGbigpl6RiYpwkSSW67z5o0yZDCfDEiXDqqdCjhwmwpPXmSrAkqVx++QV23hlG
   jYL69dN8s7lz4cADw0jka65J880k5Sq7Q0iS0q5DB5g9G7p0SfONCgvhtNOgdm1boUkqlt0hJElp
   NWsWPPYYDB+egZvdemtYdv7f/0yAJZWbSbAkab09/DC0agWNG6f5Rs89F2qAhwyBatXSfDNJ+cBy
   CEnSevn9d9h+e/j0U2jSJI03+uILOPpoePttN8JJKhW7Q0iS0ubxx+GII9KcAE+fHkYid+liAiwp
   pVwJliSV2eLFYRW4f3/Ye+803WTJEjjySDj4YLj99jTdRFJF5EqwJCkt+vaFHXdMYwIMoe3ExhuH
   DXGSlGJujJMklUkcw733hgEZafPcc2Es8vDhULlyGm8kKV+ZBEuSyuTNN6FSpbBXLS1Gj4Yrrggj
   kWvXTtNNJOU7yyEkSWVy771w/fVpatX7++9wyilwzz1prrWQlO/cGCdJKrVhw+DUU2HcOKhSJcUX
   j2No3Rrq1IFu3VJ8cUn5xIlxkqSUuvdeaNcuDQkwwP33w6RJ0Lt3Gi4uSatzJViSVCrjx0Pz5jBh
   AtSokeKLf/ppKIMYMgS22SbFF5eUb2yRJklKmQcfhIsuSkMCPGsWnHlmmL5hAiwpQ1wJliSV6Ndf
   Ybvt4JtvoH79FF44jsNEuO22C1m2JKWANcGSpJR44gk44YQUJ8AAnTvDTz+FvsCSlEGuBEuSirV0
   aRiR/MILsO++KbzwsGFw3HEweHBYCZakFLEmWJJUbi+/DA0bpjgBnj0bTj8d/vtfE2BJiXAlWJJU
   rIMPhquugr/8JUUXjGM444zQD7hLlxRdVJJWsSZYklQuI0bAxImhe1nKPPYYjBkDn3+ewotKUtmY
   BEuS1uk//4G2bVM4HOPbb+Gmm+CTT2DDDVN0UUkqO8shJElrNX067LxzGJFcu3YKLrhoEey/P1x+
   OVxySQouKElr58Y4SdJ669o11AGnJAEG+Mc/wjCMiy9O0QUlaf1ZDiFJ+oNFi0IS/O67KbrgwIHQ
   uzeMHAlRsYszkpQRrgRLkv6gXz/YbTfYddcUXGzWLDjvPHjySahbNwUXlKTyMwmWJP3BI4/A1Ven
   4EJxDJddBqeeCscck4ILSlJqWA4hSVrNsGFhU9xxx6XgYj17ho4QTz+dgotJUuqYBEuSVvPoo3Dp
   pVC5cjkvNH48XH89vPcebLRRSmKTpFSxRZokaaVffw1TjL/7DrbYohwXWrYMDjkETjsNrr02ZfFJ
   UmnYIk2SVCZPPx3KIMqVAAPcfz9UrQrXXJOSuCQp1VwJliQBUFgYhmM8+SQcdFA5LvTNN1BQAEOH
   QuPGKYpOkkrPlWBJUqkNHBgmGR94YDkusmRJaId2xx0mwJKymkmwJAkIG+KuuKKcsyz+/e/QC9ip
   cJKynOUQkiSmTIE99oAff4QaNdbzIl9+CUcfDSNGQMOGKY1PksrCcghJUqk8/jiceWY5EuDFi0MZ
   xL33mgBLygmuBEtSnluyBLbZBt5+O4xKXi833wxffQX9+5eznkKSyq80K8EOy5CkPNe/PzRpUo4E
   eOjQsJQ8cqQJsKScYTmEJOW5Ll3g8svX85MXLgxlEP/5D2y5ZUrjkqR0shxCkvLY99/DwQfD5Mlh
   tkWZtW8fxiP36+cqsKSsYTmEJKlYTzwRFnLXKwEeNgx69ICvvzYBlpRzTIIlKU8tXhzGJH/00Xp8
   8pIlcOGFYTxyuWcsS1LmWRMsSXlqwIAwJnnHHdfjk++5Bxo0gLPPTnlckpQJrgRLUp56/PH1HOw2
   Zgw8+CAMH24ZhKSc5cY4ScpDEyfCvvuGDXEbbVSGTywsDDvpzjoLrrwyXeFJUrk4MU6StFbdu4c8
   tkwJMIR+alFUjp5qkpQdXAmWpDyzdCk0bgxvvAG7716GT/zxR9hnH/j441BMLElZypVgSdIfvPEG
   NGxYxgQ4juHSS+G660yAJVUIJsGSlGfWa0Ncr17w889www1piUmSMs1yCEnKI1Onwm67hQ1xm2xS
   yk/65ZewbPzGG7D33mmNT5JSwXIISdJqevSAv/ylDAkwhBKIc881AZZUodgnWJLyRGFh6ArRr18Z
   Pumdd+CTT2DUqLTFJUlJcCVYkvLEe+9BzZqhwUOpLFgQWqH9979QvXpaY5OkTDMJlqQ88fjjcNFF
   ZRjydscd0KwZHHdcWuOSpCS4MU6S8sCMGbDDDmFS3GableITvv0WDj0URo6E+vXTHZ4kpZQb4yRJ
   APTuDSecUMoEuLAw9ATu1MkEWFKFZRIsSXngqafg/PPLcPLixXDZZWmNSZKSZHcISargvvgCfv0V
   DjusFCf/8gt06ABvvw2VK6c9NklKiivBklTB9egB550HlUrzL/7114eewHvtle6wJClRrgRLUgW2
   eDH06QOff16Kk997Dz74AL75Ju1xSVLSXAmWpArslVdg111hu+1KOHHhwlU9gcs0Tk6ScpNJsCRV
   YKXeEHfXXbD77qGFhCTlAfsES1IFNW0a7LILTJlSwsC3sWNh//3hyy+hYcOMxSdJ6WKfYEnKY716
   wamnlpAAxzFcfTW0b28CLCmvmARLUgUUx6UshXj55TBG7pprMhGWJGWNEpPgKIpaRlE0JoqiH6Io
   ar+OcwqiKPoiiqJRURQNSnmUkqQy+fxzWLYMDjywmJPmzYN27cJmuKpVMxabJGWDYlukRVFUGXgE
   OBKYCgyNomhAHMeji5yzGfBf4Jg4jqdEUVQ3nQFLkkr21FPQpg1ExVXE3XlnyJJLNUVDkiqWkvoE
   NwfGxnE8ESCKor5AK2B0kXPOAl6I43gKQBzHM9IQpySplObPh+eeg6++Kuak776Dbt1KOEmSKq6S
   yiEaAJOLvJ6y/L2imgC1oyh6P4qiYVEUnZPKACVJZfPSS9C8eTH73OIYrroKbroJ6tfPaGySlC1K
   WgkuTU+zKsDewBHAxsBnURQNjuP4hzVP7NSp08rnBQUFFBQUlDpQSVLp9OgBF11UzAkvvBD6p7Vt
   m6mQJCmtBg0axKBBg8r0OcX2CY6iaH+gUxzHLZe/7gAUxnF8d5Fz2gMbxXHcafnrJ4A34zh+fo1r
   2SdYktJs0iTYe+/QG3jDDddywty50LQp9O4NhxyS8fgkKRNS0Sd4GNAkiqLGURRVBU4HBqxxTn/g
   oCiKKkeGw1sCAAAgAElEQVRRtDHQAvh2fYOWJK2/p5+G009fRwIM8K9/hY1wJsCS8lyx5RBxHC+N
   oqgt8BZQGegex/HoKIouXf7xbnEcj4mi6E3gK6AQeDyOY5NgScqwOA5JcJ8+6zhh9Gh48kn4+uuM
   xiVJ2cixyZJUQXz6KVx4IXz77Vpao8UxHHEEnHxymBAnSRWYY5MlKY/06gXnnLOO3sDPPguzZsEV
   V2Q8LknKRq4ES1IFsGhR6Hb2xRew9dZrfHDOnLAZrl8/OOCAROKTpExyJViS8sRrr8Eee6wlAYaw
   Ge7II02AJamIkvoES5JyQM+ecO65a/nA99+HzXCjRmU8JknKZpZDSFKOmzEDdtgh9AjedNM1Pnj8
   8aEl2vXXJxKbJCWhNOUQrgRLUo579lk47ri1JMCvvw5jx4Y5ypKk1VgTLEk5bkVXiNUsXgzXXgsP
   PghVqyYSlyRlM5NgScph330HEyfCUUet8YHOnUONxHHHJRGWJGU9yyEkKYc98wycdRZsUPRf859/
   hn//O0zPkCStlRvjJClHFRbCdtuFkt9mzYp84IILoE4duPfexGKTpCS5MU6SKrCPP4YaNWCvvYq8
   OWQIvPkmjBmTWFySlAusCZakHNWz5xpjkgsL4eqr4Y471tIqQpJUlCvBkpSDFiyAF1+Er78u8uYz
   z4RE+LzzEotLknKFSbAk5aABA2CffaBBg+VvzJkDHTqEzLiSv+STpJL4L6Uk5aBevdYYk3zHHXDk
   kdCiRWIxSVIusTuEJOWY6dNhp51gyhTYZBPCVLj99w+1EVttlXR4kpS40nSHcCVYknJM375w0knL
   E2CA666DG24wAZakMjAJlqQcs6IrBABvvQWjR0O7donGJEm5xiRYknLIN9+EgXCHHw4sXgzXXAMP
   PADVqiUdmiTlFJNgScohvXrB2WdD5crAI49A48ZwwglJhyVJOceNcZKUI5YtCznvG2/AbptPh912
   g48+gp13Tjo0Scoqjk2WpApk0CCoWzfkvlx0U+iRZgIsSevFJFiScsTK3sDDh8Nrr8GYMUmHJEk5
   y3IIScoB8+aF6XBjRsdsedpBcMEFcOGFSYclSVnJPsGSVEG8/DIccABsObAPLFoE55+fdEiSlNMs
   h5CkHNCzJ7Q5YyG0bw/9+kEl1zAkqTz8V1SSstxPP8GQIdBq9L+hoCAsCUuSysWVYEnKcn36wClH
   /s7G3R+Gr75KOhxJqhBcCZakLNerF5w77R64/vqwO06SVG52h5CkLDZyJJx09AImVN+dSt+Ogg03
   TDokScp6doeQpBzX6+ll/HVZTyo9eL8JsCSlkCvBkpSlli6FRnXm8f5uV7Pzx09AVOyihiRpOccm
   S1IOe++F32g0fzw7P/43E2BJSjHLISQpS/W8+TvOOXAC7LJL0qFIUoVjOYQkZaE5H4+k0SGN+eF7
   2HyHmkmHI0k5xY1xkpSL4pgXLniNQ3b/zQRYktLEmmBJyjZ9+9Jr2pFc3r1h0pFIUoVlOYQkZZN5
   85i8w2HsNf8Tpk6vYlc0SVoPdoeQpFzz73/Tu961nNbCBFiS0smVYEnKFhMmEO+zL7vW/ZnHnqzC
   QQclHZAk5SY3xklSLrn+ekacfjcLl1bhwAOTDkaSKjaTYEnKBgMHwogR9KzUhnPOcTaGJKWb5RCS
   lLSlS6FZM5b84zYatD2FTz+FHXZIOihJyl2WQ0hSLujaFbbYgrc2OpkmTUyAJSkT7A4hSUmaMQNu
   uw0GDqTnbRHnnJN0QJKUHyyHkKQkXXYZVKvGb7f+h222gQkToHbtpIOSpNxmn2BJymZffAEvvwyj
   R/Pcc3DUUSbAkpQp1gRLUhLiGK6+OpRC1KpFz55YCiFJGWQSLElJ6NsX5s2DCy9k/HgYMwaOPTbp
   oCQpf1gOIUmZNm8e3HhjSIQrV+aZZ+CMM6Bq1aQDk6T8YRIsSZl2111wyCFw4IHEMfTsCX36JB2U
   JOUXk2BJyqRx40Jf4JEjARg8GDbYAPbbL+G4JCnPWBMsSZn0t7+Fo0EDIKwCn3uuY5IlKdNcCZak
   THn7bRg1KtQCA4sWQb9+MGJEwnFJUh5yJViSMmHJErjmGnjgAdhwQwBeew322AO22Sbh2CQpD5kE
   S1ImPPJIyHZPPHHlWytKISRJmefYZElKt+nTYbfd4KOPYOedAZgxA7bfHiZPhk03TTg+SapgSjM2
   2ZVgSUq3jh3Dku/yBBjg2Wfh+ONNgCUpKW6Mk6R0GjoUXn89jIQromdPuPXWhGKSJLkSLElpU1gI
   V18Nd94JNWuufPu772DSJDjyyARjk6Q8ZxIsSenyzDOwbBmcd95qb/fqBWedFYZkSJKS4cY4SUqH
   OXNCDfCLL0KLFivfLiyEbbeFAQNgzz0TjE+SKjA3xklSUm6/HY46arUEGEKDiJo1TYAlKWn+Mk6S
   Um3MGOjeHb7++g8fsjewJGUHyyEkKZXiGI4+OvQ/a9dutQ/Nnw8NGsA330D9+gnFJ0l5wHIIScq0
   55+Hn3+Gtm3/8KEBA6B5cxNgScoGlkNIUqrMnQvXXQe9e6+19UOPHpZCSFK2sBxCklLl73+HqVND
   D7Q1TJ0Ku+8OU6bAxhsnEJsk5ZGUlENEUdQyiqIxURT9EEVR+2LO2y+KoqVRFJ26PsFKUk4bMwae
   eALuuWetH+7VC1q3NgGWpGxRbBIcRVFl4BGgJbALcGYURU3Xcd7dwJtAsVm3JFU4cQxXXQU33wxb
   bbXWD/foAW3aZDwySdI6lLQS3BwYG8fxxDiOlwB9gVZrOe8q4Hng/1IcnyRlv+efh+nT17oZDmDw
   4JAI779/huOSJK1TSUlwA2BykddTlr+3UhRFDQiJ8aPL37LwV1L+WLEZ7pFH1jkHecUqcOTvySQp
   a5TUHaI0Ce1DwN/jOI6jKIoophyiU6dOK58XFBRQUFBQistLUha7/XYoKIBDDlnrhxcsgOeeW+vc
   DElSigwaNIhBgwaV6XOK7Q4RRdH+QKc4jlsuf90BKIzj+O4i54xnVeJbF5gPXBzH8YA1rmV3CEkV
   y5gxcPDBIcPdcsu1ntKnT5gS9+abGY5NkvJYabpDlLQSPAxoEkVRY+An4HTgzKInxHG8XZEbPgW8
   smYCLEkVzorNcDfdtM4EGEIpxAUXZC4sSVLpFJsEx3G8NIqitsBbQGWgexzHo6MounT5x7tlIEZJ
   yj4lbIYDmDwZhg+H/v0zGJckqVQcliFJZTV3LjRtGmodDj54nafdeWdIhB99dJ2nSJLSoDTlECbB
   klRW7dvDTz+tdTLcCnEMO+4IzzwDLVpkMDZJUkpqgiVJRX39NTz5JIwaVexpn34aOqY1b56huCRJ
   ZVLi2GRJ0nKFhXDZZfCvf0G9esWe2qMHnH++vYElKVu5EixJpdW9e0iEL7mk2NPmz4cXXihxsViS
   lCCTYEkqjV9+Ce3Q3n0XKhX/S7QXXwwjkuvXz1BskqQysxxCkkrj+uvhvPNgjz1KPHXFmGRJUvZy
   JViSSjJwIHzwAXzzTYmn/vgjfPEFnHRSBuKSJK03V4IlqTiLFsHll8PDD8Mmm5R4+lNPwVlnwYYb
   ZiA2SdJ6cyVYkopz992wyy6lWtpdtix0T3vllQzEJUkqF5NgSVqXH36Azp1DfUMpvPtu6Jy2555p
   jkuSVG6WQ0jS2sRxKIPo2BEaNSrVpzzxBFx0UZrjkiSlhGOTJWlteveGe++FYcPC6LcS/PIL7LRT
   2Bi36aYZiE+StE6OTZak9TFrVmiJ9vLLpUqAAXr1gpNPNgGWpFzhSrAkren886FGjVAPXApxHPbO
   PfEEHHhgmmOTJJXIlWBJKqt33w19gcsw8/jTT0MifMABaYxLkpRSboyTpBXmzYNLLoGuXcNKcCmt
   2BAXFbvmIEnKJpZDSNIKf/sb/Pxz2BRXSrNnwzbbwPffwxZbpDE2SVKpWQ4hSaU1dCg880yZyiAA
   +vaFI480AZakXGM5hCQtWRLqGe6/HzbfvEyfam9gScpNJsGSdO+9UL8+nH12mT7tyy9h+nQ46qg0
   xSVJShvLISTlt+++gwcegOHDy7yzrXt3uOACqFw5TbFJktLGjXGS8ldhIRQUwGmnwdVXl+lTFyyA
   hg1hxIiwMU6SlD1KszHOcghJ+euxx2DxYrjyyjJ/6vPPw377mQBLUq6yHEJSfpo4EW6+GT74YL3q
   GR59FNq3T31YkqTMcCVYUv6J49DS4W9/g113LfOnjxwJkyfD8cenITZJUkaYBEvKP926we+/ww03
   rNend+0KF18MG/i7NEnKWW6Mk5RfJk6EffeFDz+EXXYp86fPmRPqgL/+Gho0SH14kqTyc2OcJBVV
   WAgXXhhWgNcjAYYwUfmww0yAJSnXmQRLyh/dusHcuaEWeD3EcSiFuOyyFMclSco4K9ok5YcJE+Af
   /4CPPlrvYt7Bg0MOfcQRKY5NkpRxrgRLqvhWlEHceCM0bbrel1mxClzJfzklKee5MU5SxdelC/Ts
   CZ98st4zjmfNgu22g7FjoW7dFMcnSUqp0myMsxxCUsX2/fdwyy3w8cfrnQAD9OgBJ55oAixJFYVJ
   sKSKa8kSOOcc6NQJdt55vS+zYkPcU0+lLjRJUrKsbJNUcd15J9SqBVdeWa7LDBwIG24IBxyQorgk
   SYlzJVhSxfT556EW+IsvICq2LKxEXbqEDXHlvIwkKYu4MU5SxTNvHjRrBnfcAa1bl+tSP/4Ie+8d
   HjfZJEXxSZLSqjQb40yCJVU8V1wRGvr27FnuS/3977BoETz4YArikiRlhN0hJOWf118Px8iR5b7U
   ggXQvTt89lkK4pIkZRWTYEkVx4wZcPHF0KcP1KxZ7sv16QMtWsAOO6QgNklSVrE7hKSKIY5DAnzW
   WXDooSm5XOfOcPXVKYhNkpR1XAmWVDE8+mjYvda3b0ou99FHsHgxHHVUSi4nScoyJsGSct9XX8E/
   /xnGIlerlpJLdu4MbdvaFk2SKiq7Q0jKbfPmwX77hTYO556bkktOmhQ6rE2cCDVqpOSSkqQMskWa
   pIrv4oth4ULo1Stll+zQIXSGeOihlF1SkpRBtkiTVLE9+ywMGgQjRqTskgsWwBNP2BZNkio6k2BJ
   uWnCBLjqKnjjjZTWLPTubVs0ScoHJsGScs+SJXDmmaEOeJ99UnbZwkK4/37o0iVll5QkZSn7BEvK
   PbfcArVrQ7t2Kb3sG2/ARhtBQUFKLytJykKuBEvKLa++Cs88A8OHQ6XU/hx/331w/fW2RZOkfGAS
   LCl3TJgAF14IL74IW2yR0kuPGAHjxkHr1im9rCQpS1kOISk3LFoUMtS//x0OPDDll7///jAiuUqV
   lF9akpSF7BMsKTdccQX88gs891zK6xVWDMcYPx5q1kzppSVJCbBPsKSKoXdvePddGDYsLQW7nTtD
   mzYmwJKUT1wJlpTdvvkmtGt47z3YY4+UX372bNhuO/jiC9h665RfXpKUgNKsBFsTLCl7zZkDp50W
   2jakIQGGMB3umGNMgCUp37gSLCk7FRbCn/8cukB065aWWyxeHCbDvfRSSmduSJISZk2wpNx1++3w
   f/8Hzz6btlv06QM77WQCLEn5yCRYUvbp3x8efxyGDoWqVdNyi2XL4N//hkcfTcvlJUlZziRYUnb5
   9lu46CJ47TXYcsu03eall2CzzRyRLEn5yo1xkrLHb7/BySfDPfdA8+Zpu00cw513QseOjkiWpHxl
   EiwpOyxbBmedBS1bwvnnp/VWb78NS5bACSek9TaSpCxmEiwpO3TsCPPnh/nFaXbnndChA1TyX0BJ
   ylvWBEtKXvfu8MILMHgwVKmS1lt9/DFMmQJ/+UtabyNJynImwZKS9f77YRX4ww+hbt203+6uu6B9
   e9jAf/0kKa/5vwFJyfn+ezjjDPjf/0LD3jQbNgxGjoQXX0z7rSRJWc6KOEnJmDkTjj8e7rgDDj88
   I7fs1CnUAlerlpHbSZKymGOTJWXeokVw9NHQokVoh5YBQ4aEKcxjx5oES1JFV5qxyaVaCY6iqGUU
   RWOiKPohiqL2a/n42VEUjYyi6Ksoij6JomiP9Q1aUgUXx3DppVC7dhjZliGdOoXSYxNgSRKUoiY4
   iqLKwCPAkcBUYGgURQPiOB5d5LTxwCFxHM+Ooqgl8BiwfzoClpTjbr4ZRo+GgQMz1qNs8GAYNSpM
   iZMkCUq3Ma45MDaO44kAURT1BVoBK5PgOI4/K3L+50DDFMYoqaL473/huefgk0+gevWM3dZVYEnS
   mkqTBDcAJhd5PQVoUcz5FwKvlycoSRXQ88+HKRUffwybb56x2372WVh4HjAgY7eUJOWA0iTBpd7N
   FkXRYcAFwIHrHZGkiueDD+CKK+Ctt2DbbTN663/+E266CapWzehtJUlZrjRJ8FSgUZHXjQirwatZ
   vhnucaBlHMe/ru1CnTp1Wvm8oKCAgoKCMoQqKSd9/TW0bh16ATdrltFbf/AB/PADtGmT0dtKkjJs
   0KBBDBo0qEyfU2KLtCiKNgC+A44AfgKGAGcW3RgXRdHWwEDgr3EcD17HdWyRJuWbH3+Egw6Ce+8N
   QzEyKI7hT3+Ctm3hr3/N6K0lSQkrTYu0EleC4zheGkVRW+AtoDLQPY7j0VEUXbr8492AW4BawKNR
   FAEsieO4eXm/AEk57Kef4Igj4IYbMp4AA7z8MixcCGedlfFbS5JygMMyJKXejBlw6KFhCbZDh4zf
   fulS2H13eOABOPbYjN9ekpSwlA3LkKRS++23MA3u5JMTSYABnn4a6tWDli0Tub0kKQe4EiwpdebO
   hWOOgX33hYcegqjYH8LTYsEC2HHH0I54f0f2SFJeciVYUuYsXAitWsHOO8ODDyaSAAM88gjst58J
   sCSpeK4ESyq/hQvh1FNh002hd2+oXDmRMGbNCjn4Bx9A06aJhCBJygKlWQk2CZZUPgsWwCmnQM2a
   8MwzUKVKYqFccw0sWQJduiQWgiQpC6SkRZokrdP8+WEDXN260LMnbJDcPymjR0OfPvDtt4mFIEnK
   IdYES1o/8+bBiSeGNgwJJ8AA118fmlFsvnmiYUiScoQrwZLKbt48OOEE2GYb6N49sRrgFd58M4xH
   fumlRMOQJOUQV4Illc3s2WECxbbbZkUCvGQJXHcd3H8/VK2aaCiSpBxiEiyp9H75BQ47DHbbDZ54
   IvEEGKBbN2jYMCxMS5JUWnaHkFQ6P/4YJsGdfjrcemtifYCLmjkztEIbODDk5ZIkgS3SJKXK6NFh
   Etx110G7dklHs9JFF8Emm4ThdJIkrWCLNEnlN3Ro6AJxzz1w7rlJR7PSJ5+EDXG2RJMkrQ+TYEnr
   9vbbcPbZYQPcSSclHc1KS5bAZZfBAw+EIXWSJJWVG+MkrV337nDOOfDii1mVAAP85z/QoAG0bp10
   JJKkXGVNsKTVxTH84x/Qty+8/jrsuGPSEa1m0iTYe28YPBh22CHpaCRJ2ciaYElls2gRnH8+TJwI
   n32WlePXrrkGrr7aBFiSVD4mwZKCmTPhlFPCGOT33oONNko6oj948cWwEa5v36QjkSTlOmuCJcE3
   38D++4fj2WezMgGeORPatoUnn4Rq1ZKORpKU66wJlvLdyy/DJZfAffdlVQu0NZ1zDtSpY09gSVLJ
   rAmWtG6FhXDbbaELxGuvwX77JR3ROr36Knz6KXz1VdKRSJIqCpNgKR/NmRNWfX/5JQzD2HLLpCNa
   p99+g8svh169oHr1pKORJFUU1gRL+WbMmFD7u/nmMHBgVifAELpBnHgiFBQkHYkkqSJxJVjKJ888
   A9deC3feCRddBFGx5VKJ69cv9AMeMSLpSCRJFY1JsJQPFiwIzXU/+ADefRf23DPpiEo0ZUroBvH6
   65ZBSJJSz3IIqaL77jto0QLmzoXhw3MiAS4shPPOC6UQ++6bdDSSpIrIJFiqqOI47CY76CC44gro
   0wdq1Eg6qlJ54IEwvO7vf086EklSRWU5hFQRzZoFl10WhmC88w7stVfSEZXa55/DPfeEx8qVk45G
   klRRuRIsVTTvvAN77AH168OwYTmVAM+aBaefDo89Bttum3Q0kqSKzIlxUkWxYAF06AAvvBBmCx91
   VNIRlUlhIbRqBU2ahHIISZLWlxPjpHzx2Wdw4YWw++4wciTUrp10RGV2//0wY0bI4SVJSjeTYCmX
   zZ0LN98Mzz4LnTvDaadlfe/ftRk0KCTBQ4ZA1apJRyNJygfWBEu56p13wsrvr7/CqFHQunVOJsAT
   JsAZZ0Dv3rD11klHI0nKF64ES7lm5ky48cYw9KJrVzj22KQjWm9z58LJJ0PHjnDEEUlHI0nKJ64E
   S7misDC0TdhllzBCbdSonE6ACwuhTRvYZx+46qqko5Ek5RtXgqVcMHQoXHklVKkCb72VU23P1qVT
   J5g6NZRB5GAVhyQpx7kSLGWzmTPD0IsTTwxT3z76qEIkwN27h+T35ZehWrWko5Ek5SOTYCkbLVoU
   2iXsvDNssAGMHh1qByrl/l/ZN96Am24Kj/XqJR2NJClfWQ4hZZM4hn79wtCL3XaDDz+Epk2Tjipl
   hg+Hc8+F/v1hxx2TjkaSlM9MgqVs8ckn8Le/weLF8MQTcPjhSUeUUt99F6o6unWDAw5IOhpJUr7L
   /d+tSrlu6NDQ5eGss0Ld77BhFS4BnjAhTHG+4w449dSko5EkySRYSs6XX0KrVnDKKWGJ9PvvQ61A
   Baj7LWrqVDjySGjfHs4/P+loJEkKKtb/baVc8M03YbzxsceGFd+xY8MKcAVsk/DzzyEBvvTS0OFN
   kqRsYRIsZcpnn4WV38MPhxYtYNw4uOYa2HDDpCNLi8mT4ZBD4Oyzw4A7SZKyiRvjpHSK49AL7N//
   DlnhDTdA376w0UZJR5ZW48eHMchXXQXXXZd0NJIk/ZFJsJQOixfDs8/CvfeGGt/27aF169Dzt4Ib
   MyZsguvYES6/POloJElau4r/f2Qpk376Cbp2hcceC31+77kHjjkmb+YCf/IJ/PnPcPfdcN55SUcj
   SdK6WRMslVccw8cfw+mnw667hlHH778P774LLVvmTQL8wguh0cXTT5sAS5KynyvB0vr69Vfo0ycM
   tpg7F9q2DSvANWsmHVlGxTE89FCY8vzWW9CsWdIRSZJUMpNgqSwKC8Mqb/fu8NprYaX37rtDH7AK
   1t+3NBYuDHW/w4eHUohttkk6IkmSSsckWCqNceOgd2/o0QNq1IALL4SHH4Y6dZKOLDGTJ4fpb9tt
   F7q/Va+edESSJJVe/i1dSaU1bVr4PX+LFvCnP8H06dCvX5j0dvXVeZ0Av/suNG8eGl707WsCLEnK
   PVEcx5m5URTFmbqXtN5+/RVefDHU+o4YASedBGedFZre5kF7s5IsWQI33wzPPAM9e4Y/FkmSsk0U
   RcRxXOzOdP+vLk2ZAv37w0svwZAhocntFVfAccdV+KEWZTFuXPh5oG7dsBi++eZJRyRJ0vpzJVj5
   J45h9Gh4+eWQ+I4fDyecACefDEcf7e/211BYCI88ArfdFlaBr7kmb7q+SZJylCvB0gpz5sDAgfDm
   m6GP19KlIem9+244+GCoUiXpCLPSd9+FPYAAn34KO+6YbDySJKWKK8GqmAoLw+/s33orJL4jRsD+
   +4eWZi1bwi67uJxZjHnz4N//hi5d4J//DC2Q87ADnCQpR7kSrPxRWAjffAODBsEHH8CHH0KtWiHh
   bd8eDj3UModSiOPQAOOGG+Cgg2DkSGjYMOmoJElKPVeClZuWLYOvvw4J76BB8NFHIek99NBVx9Zb
   Jx1lTvnsM+jQAX77LbRAPvjgpCOSJGn9lGYl2CRYueHnn+Hzz2Hw4HAMGwYNGsAhh0BBQUh6GzRI
   OsqcNGIE/OMfMGoU3HILtGkDlSsnHZUkSevPJFi5afbs8Hv44cNXJb6//x6GVrRoEWp7mzeH2rWT
   jjSnDRkS9gUOHgwdO8JFF0G1aklHJUlS+ZkEK7vFMUydGjawffFFePzyyzCZbffdoVmzVUlvkybu
   zEqBZcvglVfg/vvD2ON27eCSS2DjjZOOTJKk1DEJVvb4v/+Db78Nx+jR4XHkyNChoVmzcOy1Vzia
   NPH38Sk2fXqY8PbYY6F0+vrr4dRTHYInSaqYTIKVWUuXhuXFH35YleiueCwshKZNQ2uyFY977glb
   bmmrsjRZuhTeeQe6d4f33oNTTgklD3/6k3/kkqSKzSRYqbdgQZiwNm5cOMaOXfV88mTYYgvYYYfV
   k92mTaFePTOvDFi6NDTL6NcvDMPbbjs4/3w480yoWTPp6CRJygyTYJVNHIeyhcmTVx2TJq16PnEi
   zJwJjRvD9tuvOnbYITw2bgwbbpjwF5F/fvstrPS++Sb07x++Da1bw2mnwbbbJh2dJEmZZxKsII5D
   d4Wff151TJu26nFFkjtlStgh1ahR6LHbqNHqz7fZJrQhs143UYsXh7ZmKxLfL7+EAw+EY44Jk6BN
   fCVJ+c4kuCKbPz+sys6YER5XPJ8xI+yCKpro/vwzVKkS6m/XdqxIdhs1cqpaFvrll9At7pNP4OOP
   Q4vkJk1Ca+Rjjw2tkjfaKOkoJUnKHilJgqMoagk8BFQGnojj+O61nNMZOBaYD7SJ4/iLtZxjErym
   hQtDT9x1HbNmrZ7gFn0ex1C3LtSps+pxxfM1k9x69WCTTZL+alWCJUtgwoQwtOKLL1Yd8+fD3nuH
   1d6DDgod4zbdNOloJUnKXuVOgqMoqgx8BxwJTAWGAmfGcTy6yDnHAW3jOD4uiqIWwH/iON5/LdfK
   7SQ4jsOmsHnzYO7c8Fia53PmrDvJhbBbaV1HrVp/THBXPN9444xuNBs0aBAFBQUZu19FNX/+quqT
   8ePh++/hu+/C448/hmqTXXZZ1TWuWbNQhVKeb7Xfu9zm9y93+b3LbX7/cltpkuCSuoQ2B8bGcTxx
   +QX7Aq2A0UXOOQl4GiCO48+jKNosiqJ6cRxPX+/I12bZslAMuWTJuh8XLvzjsWBB2d4v+rH581cl
   s8ht4MMAAAWcSURBVPPnQ9WqoVygevWwsrrm86Lv1agRVmE32WTtCe5mm+XUJjL/MVi3JUtWVaL8
   3/+FY8Xz6dNX32e4YAE0bBgqT7bdFnbaKazu7rRT6OSQjoltfu9ym9+/3OX3Lrf5/av4SkqCGwCT
   i7yeArQoxTkNgT8mwa1aFZ/EFvcIIQmtWjXUt675WKVKSCo32ig8rnkUfX+zzUKCuraPFT023nj1
   xNYNYTkjjkNr4mXLQtuwFT9Drfg5Z9GiP/7ss+Z78+aF/YRz5qz7cfbs8HPSioX6zTdf9bj55rDr
   rtCy5aqS67p17RQnSVI2KCkJLm39wpr/W1/r550wtVsYfVspgqgScVQJNoxgo0rh/SgK71WKVr6m
   UiViKv0hc1hbZcVq7y0MR4nnJfxetsWzrvemToXXX8/sfSEksiuS2KIJ7dper/mxKAoT0SpXDke1
   auFY82edNd9bcd4mm4Ta20aNwmONGn98XLGo70RnSZJyS0k1wfsDneI4brn8dQegsOjmuCiKugKD
   4jjuu/z1GP6/vbsLlaKOwzj+fagkIorqoigPZGBgipFUSNF7hkjkZQkR1U1YmkT0okK3XRRl1E2U
   BlERcRJRkOpAdVdhddL0GL1cRBaiBAUFoXGeLmYSX86+udvMTvt8rs7u/GEemN0zv/3PzP8HNxx/
   O4SkBt8QHBERERFN0u89wZ8DcyVdDPwC3AmsOG7MVmAV8HZZNP820/3AnYJERERERFSlbRFs+29J
   q4D3KZZI22h7r6QHyu0v294uaZmk74E/gfv+89QREREREX2orFlGRERERMSwqPRxHkmrJe2VtFvS
   CU03YvhJelTStKRz684S3ZP0TPnd2ylps6Sz684U7UlaKukbSd9JeqLuPNE9SWOSPpK0pzzfPVx3
   puiNpFMkTUraVneW6E25VO94ec6bKm/VnVFlRbCkmyjWFF5oewHwbFX7jsGQNAYsAX6sO0v07ANg
   vu3LgW+BtTXniTbKRkUvAUuBy4AVkubVmyp6cBh4xPZ8YDHwUI5f46wBpuh+lawYHi8A223PAxZy
   bG+LY1Q5E7wSeNr2YQDbByvcdwzGc8DjdYeI3tmesD1dvvyMYi3vGF5HGhWV/zP/bVQUDWB7v+2v
   yr//oDgJX1hvquiWpNnAMuBVTlwCNoZYeZXzOtuboHi2zfbvrcZXWQTPBa6X9KmkjyVdWeG+o0+S
   lgP7bO+qO0v07X5ge8dRUaeZmhBdVFOW6EO5utIVFD8+oxmeBx4DpjsNjKEzBzgo6TVJX0p6RdIZ
   rQZ3WiKtJ5ImgAtm2LS+3Nc5thdLugp4B7hkkPuP/nQ4fmuB244eXkmo6Fqb47fO9rZyzHrgkO23
   Kg0Xvcol2P8BSWcC48CackY4hpyk24EDticl3Vh3nujZqcAiYJXtHZI2AE8CT7UaPDC2l7TaJmkl
   sLkct6N8uOo8278OMkOcvFbHT9ICil9XO1V07psNfCHpatsHKowYbbT7/gFIupfiEt8tlQSKfvwM
   jB31eoxiNjgaQtJpwLvAG7a31J0nunYNcIekZcDpwFmSXrd9T825ojv7KK5a7yhfj1MUwTOq8naI
   LcDNAJIuBWalAG4G27ttn297ju05FB+yRSmAm0PSUorLe8tt/1V3nujoSKMiSbMoGhVtrTlTdEnF
   bMFGYMr2hrrzRPdsr7M9Vp7r7gI+TAHcHLb3Az+VdSbArcCeVuMHOhPcwSZgk6SvgUNAPlTNlUu1
   zfMiMAuYKGfzP7H9YL2RopVWjYpqjhXduxa4G9glabJ8b63t92rMFCcn57vmWQ28WU4g/ECbJm5p
   lhERERERI6fSZhkREREREcMgRXBEREREjJwUwRERERExclIER0RERMTISREcERERESMnRXBERERE
   jJwUwRERERExclIER0RERMTI+Qcf8kpJrd7tlwAAAABJRU5ErkJggg==
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[18]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">)</span>
   <span class="n">support</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">linspace</span><span class="p">(</span><span class="o">-</span><span class="mi">6</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">1000</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">support</span><span class="p">,</span> <span class="n">stats</span><span class="o">.</span><span class="n">logistic</span><span class="o">.</span><span class="n">pdf</span><span class="p">(</span><span class="n">support</span><span class="p">),</span> <span class="s">&#39;r-&#39;</span><span class="p">,</span> <span class="n">label</span><span class="o">=</span><span class="s">&#39;Logistic&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">support</span><span class="p">,</span> <span class="n">stats</span><span class="o">.</span><span class="n">norm</span><span class="o">.</span><span class="n">pdf</span><span class="p">(</span><span class="n">support</span><span class="p">),</span> <span class="n">label</span><span class="o">=</span><span class="s">&#39;Probit&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">legend</span><span class="p">();</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAscAAAGjCAYAAADAe7j/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzs3XmczeX7x/HXbWRfsi9jpFAooqQiDNlLUn0rSZskUql+
   Ja1C9U27lFYtvpVKSbtsE0khkuxDk7EbhMjY7t8f94yGxpxZzjmfzznn/Xw85jFzzme7zqzX3J/r
   vm5jrUVERERERKCQ1wGIiIiIiPiFkmMRERERkQxKjkVEREREMig5FhERERHJoORYRERERCSDkmMR
   ERERkQwBk2NjTCdjzDJjzEpjzKAc9jvLGHPAGHNpXo8VEREREfGDHJNjY0wcMAroBDQAehhj6h9j
   vyeAb/J6rIiIiIiIXwQaOW4GJFtrU6y1+4FxQLds9rsVGA9sycexIiIiIiK+ECg5jgdSszxem/Hc
   YcaYeFzSOzrjqcwl9wIeKyIiIiLiJ4GS49ysLf0ccK9161CbjLfcHisiIiIi4huFA2xfByRkeZyA
   GwHO6kxgnDEGoCLQ2RizP5fHYoxREi0iIiIiYWGtNTltD5QczwPqGmNqAeuBK4AeR13gpMyPjTFv
   Ap9baz8zxhQOdGyWcwQIQ/xqyJAhDBkyxOswJB/0tYsMBw/C66/Dk09CkSJw/vlw0knwzTdDOPnk
   IUydCvv2wb33wnXXQeFAv9XFF/TzF7n0tYtsGYO5Ocrx16i19oAxZgAwCYgD3rDWLjXG9M3Y/kpe
   j81D/CIiMW3pUrj6aihdGt5+G1q0+Gfbjh0wZAhYC7Nnw/33w+jR8L//QX31BRIRybeAYwzW2q+B
   r496Ltuk2Fp7faBjRUQksE8/hT594LHH4MYb4ViDHcZA8+YwbRq8+iq0agVjxkDXruGNV0QkWugG
   nBRIYmKi1yFIPulr51/vvAODBsHXX0PTptnvc/TXzxjo2xcaN4bu3eHPP6FXr9DHKvmjn7/Ipa9d
   9DNe1/saY6zXMYiI+MXEidCvnxsJrlcvf+dYutTVJj//PPznP8GNT0QkkhljAk7IU3IsIuITc+bA
   BRfkPGKcWwsXQrt2wTmXiIRGbiaHSf5ll18qORYRiRCbNsEZZ7hJdRddFJxzfvIJDBzoku6qVYNz
   ThEJnoxEzeswotKxPrdKjkVEIsChQ9ClixvhHT48uOd++GGYOROmTIFCgZZ9EpGwUnIcOgVJjvWr
   UkTEY88+Czt3ukQ22B58EPbuhVGjgn9uEZFopJFjEREPLV/u+hfPnQsnnhiaa6xcCeeeC99/n/9J
   fiISfBo5Dh2NHIuIRCBr4eab3ehuqBJjgLp13YIhN93krikiEk79+vVjeD5qxtasWUPp0qXD/g+E
   kmMREY+MHevKKQYMCP21+vWDv/6C998P/bVEJLLVqlWLqVOnBu18o0eP5oEHHsjVdadNm3b4cc2a
   Ndm1a1fYu3ooORYR8cDOnW6hj1degbi40F8vLs7VHd99N+zaFfrriUjkMsZ40mbOL2UmSo5FRDww
   YgR07BjeHsTNm0P79sHviCEi0S89PZ2BAwcSHx9PfHw8d9xxB/v27Tu8fcSIEVSvXp0aNWrw+uuv
   U6hQIVavXg3Addddx4MPPghAWloaF154IeXKlaNChQq0atUKay29evVizZo1dO3aldKlS/PUU0+R
   kpJCoUKFOHToEADbtm3j+uuvJz4+nvLly9O9e/eQvFYtHy0iEmbr1rl+xgsWhP/ajz4KDRvC7bdD
   9erhv76IRKZHH32UOXPmsHDhQgC6devG8OHDGTp0KN988w3PPvss06ZNo1atWvTp0+eIY7OORD/9
   9NMkJCSQlpYGwI8//ogxhrFjx/L999/zxhtv0LZtWwBSUlKOOE+vXr0oU6YMS5YsoWTJksyePTsk
   r1UjxyIiYTZkCNx4I9SsGf5rx8dD794aPRaJCMYE5y0I3nvvPR566CEqVqxIxYoVefjhhxk7diwA
   H374ITfccAP169enePHiPPLII8c8T5EiRdiwYQMpKSnExcXRokWLXF1/w4YNfPPNN7z88suULVuW
   woUL07Jly6C8tqMpORYRCaNVq2DCBLj3Xu9iGDQIPvwQMu54iohfWRuctyBYv349J5xwwuHHNWvW
   ZP369YBLXBMSEg5vq1GjRjYvxcVx9913U6dOHTp06EDt2rV54okncnX91NRUypcvT9myZQvyMnJF
   ybGISBg9/jj07w/lynkXQ8WKcOutMHSodzGISGSpXr36EWUOa9asIT4+HoBq1aqRmpp6eFvWj49W
   qlQpnnrqKVatWsVnn33GM888w/Tp0wFynASYkJDAtm3b2LFjRwFfSWBKjkVEwuSPP9yo8cCBXkcC
   t90Gn38Oa9Z4HYmI+NG+ffvYu3fv4bcePXowfPhw0tLSSEtLY+jQoVx99dUAXH755bz55pssW7aM
   PXv2MGzYsCPOlbUDxRdffEFycjLWWsqUKUNcXByFMta2r1KlCqtWrco2nmrVqtG5c2f69+/Pn3/+
   yf79+5kxY0ZIXruSYxGRMBkxAvr0gfLlvY7EjVz37g1PP+11JCLiR126dKFEiRKH39LT02natCmN
   GjWiUaNGNG3a9HDv4k6dOnHbbbfRpk0bTj75ZM4991wAihYtChw5IS85OZn27dtTunRpmjdvzi23
   3ELr1q0BGDx4MMOHD6dcuXI888wzh4/NNHbsWI477jjq1atHlSpVGDlyZEheu5aPFhEJg82b4ZRT
   YNkyqFLF62ic9evhtNPcEtaVKnkdjUjs8Utf32BbunQpDRs2ZN++fYdHhcNNy0eLiPjcK6/AZZf5
   JzEG18rtP/+BF17wOhIRiXQTJkwgPT2d7du3M2jQIC666CLPEuOC0sixiEiI7dsHtWrBpEmux7Cf
   LFsGrVu72uOMO6AiEibRNHLcuXNnZs+eTVxcHImJibz00ktU8XA0oCAjx0qORURC7H//g7fegilT
   vI4kex06wDXXQMbcGhEJk2hKjv1GZRUiIj5lLTz/vD86VBzLgAEqrRARyaTkWEQkhGbPhj//hC5d
   vI7k2C64wE0YnDPH60hERLyn5FhEJIReeMEtuOHneSlxcW5hkhdf9DoSERHvqeZYRCRE0tKgTh34
   /XdvV8TLja1bXazLl0Plyl5HIxIbVHMcOqo5FhHxof/9D7p29X9iDFChAlx8MbzzjteRiIh4S8mx
   iEgIWAuvvw433uh1JLnXuzeMGeNiFxEpqEKFCrF69epst7377rt07NgxzBHljpJjEZEQ+OknSE+H
   Vq28jiT3WrSAAwc0MU8k1tWqVYsSJUpQunRpqlatyvXXX8/u3buDeo2ePXsyadKkw49zSqTDTcmx
   iEgIZI4amxwr2/zFGLj+ejd6LCKxyxjDF198wa5du5g/fz7z5s1j+PDhR+xz4MCBoF/XL/XXSo5F
   RIJs1y74+GO49lqvI8m7a66Bjz6CPXu8jkRE/KB69ep07tyZ3377jUKFCvHSSy9Rt25dTjnlFABe
   e+016tatS4UKFejWrRsbNmw44vgvv/yS2rVrU6lSJe65557DCfBbb71Fy5YtAWiVcYvt9NNPp3Tp
   0nz00UdhfIX/puRYRCTIxo935RRVq3odSd7Fx8M558Ann3gdiYh4KTOJTU1N5auvvqJJkyYATJw4
   kblz57JkyRKmTZvGfffdx0cffcSGDRs44YQTuPLKK484z6effsrPP//M/PnzmThxImOyuTU1Y8YM
   AH799Vd27drFf/7znxC/upwpORYRCbL//Q969fI6ivy74QaVVoj4gTHBecsray0XX3wx5cqVo2XL
   liQmJnLfffcBMHjwYI4//niKFi3Ku+++S+/evWncuDFFihTh8ccfZ/bs2axZs+bwuQYNGsTxxx9P
   QkICAwcO5P333w/WpydkCnsdgIhINFm3DhYsgAsv9DqS/OvaFW6+GdasgZo1vY5GJHZ5VYJrjGHi
   xIm0bdv2X9sSEhIOf7xhwwaaNm16+HHJkiWpUKEC69ato2bGL4+s+9esWZP169eHMPLg0MixiEgQ
   jRsH3btDsWJeR5J/RYu61/DBB15HIiJ+Y7IMRVevXp2UlJTDj3fv3s3WrVuJj48//FzWUeQ1a9Yc
   sc2vlByLiATRu+9Cz55eR1FwV10FEXD3U0Q81KNHD958800WLlxIeno69913H+ecc87hUWOAp556
   ij///JPU1FRGjhzJFVdcke25qlSpwqpVq8IVeo6UHIuIBMmSJbBpE7Ru7XUkBdeqFWzcCMuWeR2J
   iPiFOaqA+fzzz2fYsGFceumlVK9end9//51x48YdsU+3bt0488wzadKkCRdeeCG9e/c+fK6s5xsy
   ZAjXXnst5cqVY/z48aF/MTkwXveUM8ZYr2MQEQmG+++HffvgySe9jiQ47rgDypSBRx7xOhKR6GSM
   8U1v32hzrM9txvM5TlNUciwiEgTWwkknwYQJ0Lix19EEx9y5rrxixYrIWsxEJFIoOQ6dgiTHKqsQ
   EQmCuXOhSBE4/XSvIwmepk1d0v/zz15HIiISPgGTY2NMJ2PMMmPMSmPMoGy2dzPGLDTGLDDG/GyM
   aZtlW4ox5teMbXOCHbyIiF+MHw//+U90jbAaAz16aGKeiMSWHMsqjDFxwHKgHbAOmAv0sNYuzbJP
   SWvt7oyPGwITrLV1Mh7/Dpxprd2WwzVUViEiEc1aqF3blVRE08gxwKJFrmdzSkp0Jf4ifqCyitAJ
   ZVlFMyDZWptird0PjAO6Zd0hMzHOUApIOzqOANcQEYlo8+dDXBw0auR1JMF32mmu77FKK0QkVgRK
   juOB1CyP12Y8dwRjzMXGmKXA18BtWTZZYIoxZp4xpk9BgxUR8aPx4+Gyy6JzZNUYuPRS+PhjryMR
   EQmPQMlxrsb6rbWfWmvrA12BsVk2tbDWNgE6A7cYY1rmL0wREX+y9p9642iVmRzr7q9I8GX2+9Vb
   cN8KonCA7euAhCyPE3Cjx9my1s40xhQ2xlSw1m611m7IeH6LMWYCrkxj5tHHDRky5PDHiYmJJCYm
   5voFiIh4aeFCOHgQmjTxOpLQOfNM17/5t9+gYUOvoxGJHqo3Dr2kpCSSkpLydEygCXmFcRPyzgfW
   A3P494S82sBqa601xpwBfGStrW2MKQHEWWt3GWNKAt8Cj1hrvz3qGpqQJyIR64EHXOI4YoTXkYTW
   nXe6BUGyjGWIiEScAk/Is9YeAAYAk4AlwAfW2qXGmL7GmL4Zu10KLDLGLACeB67MeL4qMNMY8wvw
   E/DF0YmxiEik+/RTuOQSr6MIvUsvhU8+8ToKEZHQ0wp5IiL5tGoVtGgB69dDoShfUunQIahRA5KS
   4OSTvY5GRCR/tEKeiEgIffYZdO0a/YkxuNfYrRtMnOh1JCIioRUDv9JFRELjs8/goou8jiJ8unaF
   zz/3OgoRkdBSWYWISD5s2wa1asHGjVCihNfRhMfevVClCqxeDRUqeB2NiEjeqaxCRCREvvoK2raN
   ncQYoFgx95q/+srrSEREQkfJsYhIPsRaSUUmlVaISLRTWYWISB6lp7vyguXL3ftYsmkT1Kvn3hcp
   4nU0IiJ5o7IKEZEQSEqCU0+NvcQY3Gs+5RSYMcPrSEREQkPJsYhIHsVqSUUmlVaISDRTciwikgfW
   KjnOTI5VESci0UjJsYhIHixa5Gpt69XzOhLvNGwIBw/C4sVeRyIiEnxKjkVE8uCrr6BzZzA5TueI
   bsaotEJEopeSYxGRPPj6a5ccx7oLL4Qvv/Q6ChGR4FMrNxGRXNqxA2rUcG3MYmnxj+z8/TdUrgxr
   1kC5cl5HIyKSO2rlJiISRFOmQIsWSowBiheHli1h6lSvIxERCS4lxyIiuaSSiiN16gTffON1FCIi
   waWyChGRXLAWEhJg2jQ4+WSvo/GHFSugbVtITY3tCYoiEjlUViEiEiSLFkHRolC3rteR+Efduq6t
   3W+/eR2JiEjwKDkWEcmFzJIKjZD+wxiVVohI9FFyLCKSC19/DV26eB2F/3TurORYRKKLao5FRALY
   udO1cNu4UZ0qjvbXX1CtGmzYAKVKeR2NiEjOVHMsIhIEU6ZA8+ZKjLNTqhQ0a+YmKoqIRAMlxyIi
   AUyeDB06eB2Ff6nuWESiiZJjEZEAJk+G9u29jsK/Ond2NdmqkBORaKDkWEQkB7//Drt3w2mneR2J
   f516KuzbB8nJXkciIlJwSo5FRHIweTK0a6cWbjkxxn2OtJS0iEQDJcciIjlQSUXutGvnPlciIpFO
   rdxERI7h4EGoXNmtjle9utfR+NuGDa68YssWiIvzOhoRkeyplZuISAHMnw9Vqyoxzo1q1dznaf58
   ryMRESkYJcciIsegkoq8adfO9YQWEYlkSo5FRI5hyhQlx3mhumMRiQaqORYRycaePVClipZFzotd
   u1x5xebNWk1QRPxJNcciIvk0YwaccYYS47woXRqaNIHvv/c6EhGR/FNyLCKSDdUb54/qjkUk0ik5
   FhHJhpLj/FHdsYhEOtUci4gcZeNGaNBAPXvzY/9+qFjRLSVdqZLX0YiIHEk1xyIi+TB1KiQmKjHO
   j+OOg9atYdo0ryMREckfJcciIkeZNg3OP9/rKCKX6o5FJJIpORYROcq0adCmjddRRK7MumNVzIlI
   JFJyLCKSRUoK/P031K/vdSSRq3592LcPVq/2OhIRkbxTciwiksX06W7U2OQ4XUNyYoz7HE6f7nUk
   IiJ5FzA5NsZ0MsYsM8asNMYMymZ7N2PMQmPMAmPMz8aYtrk9VkTEb1RSERxKjkUkUuXYys0YEwcs
   B9oB64C5QA9r7dIs+5S01u7O+LghMMFaWyc3x2Yco1ZuIuIL1kJCgkvq6tb1OprIlpzsulasXatR
   eBHxj2C0cmsGJFtrU6y1+4FxQLesO2QmxhlKAWm5PVZExE+Sk10iV6eO15FEvtq1oVAhWLnS60hE
   RPImUHIcD6Rmebw247kjGGMuNsYsBb4GbsvLsSIifpFZUqGRzoJT3bGIRKrCAbbnqt7BWvsp8Kkx
   piUw1hhTLy9BDBky5PDHiYmJJCYm5uVwEZGgmD4dOnb0Ooro0aYNTJoEfft6HYmIxKqkpCSSkpLy
   dEygmuNzgCHW2k4ZjwcDh6y1T+RwzCpcSUXd3ByrmmMR8QNroWpVmDMHTjjB62iiQ0oKnHMObNig
   0XgR8Ydg1BzPA+oaY2oZY4oAVwCfHXWR2sa4X3vGmDMArLVbc3OsiIhfLFkCJUsqMQ6mWrWgeHFY
   ujTgriIivpFjWYW19oAxZgAwCYgD3rDWLjXG9M3Y/gpwKXCNMWY/8BdwZU7Hhu6liIjkX2Z/Ywmu
   zLrjBg28jkREJHdyLKsISwAqqxARH7jkErj0UujZ0+tIosvYsTBxIowf73UkIiK5K6tQciwiMe/Q
   IahUCRYtgurVvY4muqxdC40bw+bNrrWbiIiXglFzLCIS9X79FSpWVGIcCjVqQLlysHix15GIiOSO
   kmMRiXnTpkHbtoH3k/xRv2MRiSRKjkUk5mkyXmgpORaRSKKaYxGJaQcOuJKKFSugcmWvo4lOGzbA
   aafBli2qOxYRb6nmWEQkgF9+gfh4JcahVK2a+/wuXOh1JCIigSk5FpGYNnMmtG7tdRTRT6UVIhIp
   lByLSEybMQNatvQ6iuiXmKjkWEQig2qORSRmHTrkbvf/8otrOSahs2kT1KsHaWkQF+d1NCISq1Rz
   LCKSg6VLoWxZJcbhUKWKe1u0yOtIRERypuRYRGLWjBnQqpXXUcSO1q3d51xExM+UHItIzFJyHF6t
   WsF333kdhYhIzlRzLCIxyVpXTjFjBtSu7XU0sSE1Fc44AzZvBpNjxZ+ISGio5lhE5Bh+/929P+kk
   b+OIJQkJULq0q/UWEfErJcciEpMyW7hpBDO8VHcsIn6n5FhEYpLqjb2humMR8TslxyISk5Qce6NV
   K/e511QTEfErJcciEnPWrYPt26FBA68jiT0nneRKWVat8joSEZHsKTkWkZgzc6arNy6k34BhZ8w/
   o8ciIn6kPw0iEnNUUuEtTcoTET9TciwiMWfmTCXHXtKkPBHxMyXHIhJTtm6FNWugcWOvI4ld9erB
   7t3u6yAi4jdKjkUkpnz/PZxzDhQu7HUksUt1xyLiZ0qORSSmqN7YH1R3LCJ+peRYRGKKkmN/UN2x
   iPiVsR53YjfGWK9jEJHYsGsXVKsGaWlQrJjX0cS2gwehUiVYsgSqVvU6GhGJFcYYrLUmp300ciwi
   MeOHH+DMM5UY+0FcHJx3nkorRMR/lByLSMxQSYW/aFKeiPiRkmMRiRnqb+wvrVur7lhE/Ec1xyIS
   E/buhYoVYeNGKFXK62gE4MABKF8efv8dKlTwOhoRiQWqORYRyTBnDjRooMTYTwoXhubN3Yi+iIhf
   KDkWkZigemN/Ut2xiPiNkmMRiQlKjv1J/Y5FxG9UcywiUW//flfTmpLialzFP9LT3ddm/XooU8br
   aEQk2qnmWEQEWLAAatVSYuxHRYtC06auB7WIiB8oORaRqKcWbv6mumMR8RMlxyIS9WbMgJYtvY5C
   jkXJsYj4iWqORSSqHTrk+hv/9htUr+51NJKd3buhcmVIS4Pixb2ORkSimWqORSTmLV7sJnwpMfav
   kiWhYUP46SevIxERyUVybIzpZIxZZoxZaYwZlM32nsaYhcaYX40xs4wxjbJsS8l4foExZk6wgxcR
   CUQt3CKDSitExC9yTI6NMXHAKKAT0ADoYYypf9Ruq4FW1tpGwDDg1SzbLJBorW1irW0WvLBFRHJH
   yXFkUHIsIn4RaOS4GZBsrU2x1u4HxgHdsu5grZ1trd2R8fAnoMZR58ixrkNEJFSsVXIcKVq0cGUV
   +/d7HYmIxLpAyXE8kJrl8dqM546lN/BVlscWmGKMmWeM6ZO/EEVE8mfVKihc2PU4Fn8rVw5q14b5
   872ORERiXeEA23PdRsIY0wa4AWiR5ekW1toNxphKwGRjzDJr7cx8xCkikmeZo8YmUu9fbdkCs2a5
   VhuLF8PatbB9O+zYAccdByVKuGXlateGOnXg9NNdz7oKFbyOPF8ySyvOPtvrSEQklgVKjtcBCVke
   J+BGj4+QMQnvNaCTtXZ75vPW2g0Z77cYYybgyjT+lRwPGTLk8MeJiYkkJibm+gWIiBxLRPY3XrwY
   3n0XvvkGVq+G5s2hUSPo0gVOOMENsZYpAwcPwp49sG2b22/lSnj5ZbjmGjdU3q0b9OgBDRp4/Ypy
   rVUrePttuPturyMRkWiRlJREUlJSno7Jsc+xMaYwsBw4H1gPzAF6WGuXZtmnJjANuNpa+2OW50sA
   cdbaXcaYksC3wCPW2m+Puob6HItISJx0EnzxRQTkh+npMHYsvPIKrF8PV18NF10EzZq5EeK8OHAA
   5s6F8eNh3DioUgVuu80lykWLhib+INm0CerVc/2O4+K8jkZEolFu+hwHXATEGNMZeA6IA96w1j5u
   jOkLYK19xRjzOtAdWJNxyH5rbTNjzEnAJxnPFQbetdY+ns35lRyLSNClpsIZZ8DmzT4uq9izB0aP
   hmeecaPDt98O7dsHLzM8eBAmT3bn/+03uOMOGDDA1yttnHIKfPihqxAREQm2oCTHoabkWERC4b33
   3ODpJ58E3jfsDh2C99+He+91Bbb33w9NmoT2mr/+Cg8/DD//DMOGQa9eUMh/60D16eP+T7j1Vq8j
   EZFopBXyRCRm+baF2+LFrm/Zc8+5BHn8+NAnxuAyzgkT3DVHj3bF2EuXBj4uzNTvWES8puRYRKLS
   zJk+S44PHIDHH4fERLjuOtfU97zzwh9Hixbwww/Qs6f7BA0b5mLziczkWDcURcQrKqsQkaizZQvU
   rQtbt/pkYteaNXDllVCqFLz2mus64QepqdC7N/z9t+uQUbOm1xEB7tPz7beu/lhEJJhUViEiMen7
   7+Hcc32SGH/1les6ccklMGmSfxJjgIQE1zLuggvgrLPg88+9jghQaYWIeEvJsYhEHV/UG1vrShb6
   9oWPP4b/+z9/ts0oVMhNDJwwAfr1g8ce87ymQcmxiHhJybGIRB3Pk+P0dFdX/NlnMGeOq/P1u+bN
   XR30hAlw1VWu1MIjrVq5mnERES8oORaRqLJjByxfDk2behTAtm3QoQPs2gVJSVCtmkeB5EN8/D+z
   4Tp2dJ9MD5x8ssvN//jDk8uLSIxTciwiUeWHH1z5rCeLwW3Y4FqkNW0KH30EJUt6EEQBFS/umkSf
   fjq0bg0bN4Y9BGPcp1GjxyLiBSXHIhJVPGvhtmaNu3DPnvD00z6ZDZhPhQrByJFuEuF558Hq1WEP
   QXXHIuIVJcciElU8qTdetcqNst5yC9x3X5gvHiLGwEMPuSWn27QJe4Ks5FhEvKI+xyISNf7+GypV
   gk2bwljRsHq1S4wfeMB1pohGo0fDiBGuhjpMregOHoQKFVz9eJUqYbmkiMQA9TkWkZjy009w2mlh
   TIzXrYN27eD++6M3MQbX4u2OO6BtW1i7NiyXjItzFR2qOxaRcFNyLCJRI6wlFVu2QPv2cPPN7i3a
   3XYb9O/vEuTNm8NySZVWiIgXlByLSNQIW3K8cyd06uQmrN1zTxgu6BN33eWWwb7gAvjrr5BfTsmx
   iHhBNcciEhX274fy5V3TiHLlQnyhLl1cM95Ro/y56l0oWQt9+rjyis8/h+OOC9ml9u1zdcch/5qK
   SMxQzbGIxIz586F27RAnUda62uLixV2rs1hLjMG95pdfdknxjTeGdKnpIkXg7LNh1qyQXUJE5F+U
   HItIVAhLScWjj8Kvv8L770d2H+OCKlwYPvjAtZJ45JGQXkpLSYtIuCk5FpGoEPLk+N134fXXXSlB
   JK58F2wlSsCnn8Kbb7rVAENEdcciEm6qORaRiHfwIFSsCMuWhagn7ty5rs54+nTXK07+sWABdOgA
   334LTZoE/fR//+2+tps3638SESk41RyLSEz47TeoXDlEifGmTXDppfDqq0qMs9OkiVsk5OKL3ecq
   yIoXh8aN4ccfg35qEZFsKTkWkYg3Ywa0bBmCE+/bB5ddBtdfD927h+ACUSLr5yg9PeinV2mFiIST
   kmMRiXi6Va/eAAAgAElEQVQzZrgVnIPujjtc+4uHHw7ByaPMQw9B9epusZAgU3IsIuGkmmMRiWjW
   QtWqriy4Zs0gnnjMGBgxwq1JXbZsEE8cxXbuhLPOggcegF69gnbaHTsgPh62boWiRYN2WhGJQao5
   FpGot3y5q0sNamK8cCEMGuS6MSgxzr0yZWD8eLjzTli8OGinLVsWTjkF5s0L2ilFRI5JybGIRLSg
   t3DbtQsuvxyeew7q1QviiWNEw4bw1FNuEuOuXUE7rUorRCRclByLSEQLanJsLdx8szthz55BOmkM
   uvZaN0PyppuCtoKekmMRCRclxyISsayF774L4mS8N96ARYvc0tBSMCNHusbTL78clNOddx788AMc
   OBCU04mIHJOSYxGJWCkpLlmqUycIJ/v1Vxg8GD780BUxS8EULw7jxrkuFkuWFPh0lSpBjRquHFxE
   JJSUHItIxMosqTA5zjvOhT174Mor4ZlnVGccTKecAo8/Dj16wN69BT5dq1Ywc2YQ4hIRyYGSYxGJ
   WEErqRg0yK30FsT2Y5Khd283tD94cIFP1bKl6o5FJPTU51hEIladOq7bWoFWdf76azcJb+FCOP74
   oMUmWWzbBqefDq+9Bp065fs0qanuf5gtW4Jwt0BEYpL6HItI1Fq3Dv78Exo0KMBJtmyBG2+Et99W
   YhxK5cvDO+/ADTfA5s35Pk1CgmulvHRpEGMTETmKkmMRiUgzZrjb7IXy+1vMWtdq7OqrITExmKFJ
   dtq0cS3err++QO3d1NJNREJNybGIRKQC9zceM8a1uxg6NFghSSBDh8KmTa68Ip+UHItIqKnmWEQi
   0qmnujv1Z56Zj4OTk+Hcc92MvgLVZUieLV7sRurnzIETT8zz4cnJ7vDUVNUdi0jeqeZYRKLSli2u
   5rhx43wcfPAgXHcd3H+/EmMvnHoq3H23K684dCjPh9eu7Q77/fcQxCYigpJjEYlAM2dC8+YQF5eP
   g0eNcoXKt90W9Lgkl+66C/btc1+LPDJGpRUiElpKjkUk4uS73njlShg2zNUb53smnxRYXJzrEDJ0
   KKxYkefDlRyLSCjpr4OIRJwZM/Kx+MehQ66V2IMPBmm9aSmQunXd0tLXXedKXfJAybGIhFLA5NgY
   08kYs8wYs9IYMyib7T2NMQuNMb8aY2YZYxrl9lgRkbz68083AJzniXiZt/BvvTXoMUk+DRgARYu6
   ZbvzoEED2L4d1q8PUVwiEtNyTI6NMXHAKKAT0ADoYYypf9Ruq4FW1tpGwDDg1TwcKyKSJ7NmQbNm
   UKRIHg5KTna38FVO4S+FCrmvyRNP5Km8olAhOO88V3suIhJsgf5KNAOSrbUp1tr9wDigW9YdrLWz
   rbU7Mh7+BNTI7bEiInmV55KKzHKK++93t/LFX0480ZW69OmTp+4VKq0QkVAJlBzHA6lZHq/NeO5Y
   egNf5fNYEZGAvvsuj5PxXnzRJV3qTuFfAwZAenqeFgdRciwioVI4wPZcr85hjGkD3AC0yOuxIiK5
   8ddf8NtvcPbZuTwgJQUeeQR++CGffd8kLOLi4I033OoeF1wANWoEPKRJE/jjD9i6FSpUCH2IIhI7
   AiXH64CELI8TcCPAR8iYhPca0Mlauz0vxwIMGTLk8MeJiYkkJiYGCEtEYtGPP7qkqHjxXOxsLfTv
   D3feCSefHPLYpIBOPdWNIPfvDxMnBlz+rnBht8jh999DNxXsicgxJCUlkZSUlKdjclw+2hhTGFgO
   nA+sB+YAPay1S7PsUxOYBlxtrf0xL8dm7Kflo0UkVx580FVIPPpoLnb+4AMYPhx+/jmPs/fEM/v2
   wRlnuC/0FVcE3P3RR2HbNnj66TDEJiJRocDLR1trDwADgEnAEuADa+1SY0xfY0zfjN0eAsoBo40x
   C4wxc3I6tkCvSERiWq4X/9i+He64A159VYlxJClSBF5/HQYOdPUSAajuWERCIceR47AEoJFjEcmF
   vXuhYkXYsAFKlw6wc58+rn9uPpYnFh+44w6XHL/zTo675el7QkSEIIwci4j4xZw5bvGHgEnQd9/B
   N9/AY4+FJS4JgeHDXRPjb7/NcbdixdxiMD/8EKa4RCQmKDkWkYgwfTq0aRNgp717oW9feOEFKFMm
   LHFJCJQs6Vrw9e8Pf/+d464qrRCRYFNyLCIRYfp01+krR4895roeXHxxOEKSUOrSxbUmCXAHQMmx
   iASbao5FxPdyVVu6eLHLnn/5BeK13lBUWL8eTj/dZb/162e7y19/QdWqkJbmyixERHKimmMRiQqz
   Z0PDhjkkxocOwU03wdChSoyjSfXq8NBD0K+f61udjVKlXC36nDlhjk1EopaSYxHxvYAlFa++6t73
   7ZvDThKR+veH3bvh7bePuUurVm4epohIMCg5FhHfy3Ey3ubNbnTx5ZehkH6lRZ24OHjlFbj3Xlc7
   kY3ERPc9IiISDKo5FhFf27MHKleGTZtcE4N/uf56KF9ey6RFu4EDYedOGDPmX5t27HDVNKo7FpFA
   VHMsIhHvhx/cnKxsE+NZs2DyZBgyJNxhSbgNG+a+1tm0pihb1jUp+fFHD+ISkaij5FhEfO2YJRUH
   DriJWk8/reXRYkHp0vD883DzzbBv3782t2kD06Z5EJeIRB0lxyLia0lJx0iOX3gBqlSByy8Pd0ji
   le7doXZtePLJf21q21Z1xyISHKo5FhHfyuxhu3kzlCiRZcP69dCokSurOOUUz+ITD/zxh1szeu5c
   OPHEw0/v3u3+VzpmbbqICKo5FpEIN2sWnHHGUYkxwF13ubZtSoxjzwknuK//bbcd8XTJkm5BvVmz
   PIpLRKKGkmMR8a1s642nTnWrgtx/vycxiQ/cdResXAmff37E023aqLRCRApOybGI+Na/6o3T0+GW
   W2DkyGyGkyVmFCkCo0a50eM9ew4/rUl5IhIMqjkWEV/atQuqVTuqd+3jj7vebkeNGEqMuuIKOPlk
   1+YN2LsXKlZ0Jellyngcm4j4kmqORSRizZwJZ52VJTH+4w/Xtm3kSE/jEh955hkYPdqVWOC+V5o1
   c987IiL5peRYRHzpX/XGAwfC7bcf0aFAYlx8vFtWesAAyLgD2batSitEpGCUHIuILyUlQWJixoMv
   v4TFi+Huuz2MSHzp9tth3Tr45BNAk/JEpOBUcywivrNjB9So4eqNix76260N/PLL0KGD16GJH333
   HfTqBUuWsK9IKSpUcFU45ct7HZiI+I1qjkUkIs2YAWefDUWLAv/9r1v0QYmxHEvr1u5t2DCKFIEW
   LVy+LCKSH0qORcR3pk/PKKlYuRJefBGefdbrkMTvnnwSxoyBpUtVWiEiBaLkWER8Z+pUaHe+hVtv
   hUGDXI2FSE6qVoUHH4RbbqFtG6tJeSKSb0qORcRXNm929aJNUydAaqrrUiGSG/37w7ZtNFnxAWvX
   wqZNXgckIpFIybGI+Mq0adC6xQEK/99AeOklOO44r0OSSFG4MLz0EoUH3UXLc/eTlOR1QCISiZQc
   i4ivTJkC7dK//GeSlUheNG8OHTvS9u8vVXcsIvmiVm4i4hvWwok19vH1nkTqL/3E1ZGK5NWWLfxy
   yhVcXuZrVqQU9ToaEfERtXITkYiyepVlX9pO6g29Somx5F+lSjQafjnb1/3Nmj80+CIieaPkWER8
   Y8pjc2hX8kdM/35ehyIRrlDfPpxfeg6TH/nB61BEJMIoORYRf9ixgynvb6bdgHoQF+d1NBLp4uJo
   378uU97fDNu3ex2NiEQQJcci4guHHniI6TaR8/vW8ToUiRLtbzqRqbTj0P0Peh2KiEQQJcci4r0F
   C/jl3cVUqlmc+Hivg5FoUbMmlIsvwcIPlsHPP3sdjohECCXHIuKtQ4egf3+mtH2Mdh0Lex2NRJl2
   HeOYcv7j0K8fHDzodTgiEgGUHIuIt8aMAWDqzrM4/3yPY5Go0749TN7e1C0m8/rrXocjIhFAfY5F
   xDtbt0KDBuydOIlK7RuTmgrHH+91UBJNduyAGjVgy7RFFLvgfFi8GCpV8josEfGI+hyLiL8NHgxX
   XMHsvxtz6qlKjCX4ypaF006D73c0hJ49YdAgr0MSEZ9Tciwi3vjxR/jiCxg2jClTUEmFhEz79m5Z
   ch55BCZNglmzvA5JRHxMybGIhN+BA9C/Pzz5JJQty9Sp0K6d10FJtGrfHiZPBsqUgaefdt97Bw54
   HZaI+JSSYxEJv9Gj3f3uq67izz9dGei553odlESrc86B5GRISwOuuAIqVoRRo7wOS0R8KmBybIzp
   ZIxZZoxZaYz5V7GWMaaeMWa2MWavMeauo7alGGN+NcYsMMbMCWbgIhKhNm6EoUPhxRfBGKZPd4lx
   sWJeBybR6rjjoGVLmDYNMMZ97w0fDuvXex2aiPhQjsmxMSYOGAV0AhoAPYwx9Y/abStwK/BUNqew
   QKK1tom1tlkQ4hWRSHf33XDDDdCgAeBKQDt18jgmiXqHSysA6tWDm26C//s/T2MSEX8KNHLcDEi2
   1qZYa/cD44BuWXew1m6x1s4D9h/jHDm2yxCRGJKUBDNmwINuOV9rXXLcsaO3YUn0y0yOD3cOvf9+
   +OEHmDrV07hExH8CJcfxQGqWx2sznsstC0wxxswzxvTJa3AiEkX274dbboFnn4VSpQBYscI9nTGI
   LBIy9eu777Xk5IwnSpaE556DAQNg3z5PYxMRfwmUHBd0dY4W1tomQGfgFmNMywKeT0Qi1XPPQc2a
   0L374acyR42N7i9JiBnjOqJMmZLlyW7doHZteOYZz+ISEf8pHGD7OiAhy+ME3OhxrlhrN2S832KM
   mYAr05h59H5Dhgw5/HFiYiKJiYm5vYSIRILUVHjiCdfbOEsmPGkSXHuth3FJTOnQAT76CPr1y3jC
   GBg5Epo1gx494IQTPI1PRIIvKSmJpKSkPB2T4/LRxpjCwHLgfGA9MAfoYa1dms2+Q4Bd1tqnMx6X
   AOKstbuMMSWBb4FHrLXfHnWclo8WiXaXXeaWKcvyj/DevVC5MqSkQPnynkUmMWTLFqhbFzZvhiJF
   smwYOhQWLIAJEzyLTUTCo8DLR1trDwADgEnAEuADa+1SY0xfY0zfjItUNcakAncADxhj1hhjSgFV
   gZnGmF+An4Avjk6MRSQGfPMN/PLLv5bt/f57OPVUJcYSPpUqwcknu3l4R7jnHvjtN/jqK0/iEhF/
   yXHkOCwBaORYJHrt3etGjF94ATp3PmLT3Xe7OVFZBpNFQu6hhyA93VX5HGHSJLdy3m+/QfHinsQm
   IqFX4JFjEZECeeIJaNToX4kxqIWbeKNzZ3cz4186doQmTeC//w17TCLiLxo5FpHQWLECmjeH+fNd
   l4os1q2Dhg1d7WfhQNOCRYLowAFX675oEcQf3Zg0NdUlyLNnu+JkEYk6GjkWEW9YCzffDA888K/E
   GODbb11bLSXGEm6FC7sFQSZNymZjQgIMHuzaWWjQRiRmKTkWkeAbOxZ27HALLGRDJRXipc6d4euv
   j7Hx9tth61b43//CGpOI+IfKKkQkuNLS3CS8L7+EM8/81+aDB91t7YULoUYND+KTmLdhg1uVccuW
   Y9y9mDsXunaFxYuhQoWwxycioaOyChEJv3vugSuvzDYxBpg3D6pWVWIs3qlWDWrVgp9+OsYOZ50F
   V1zhWqqISMxRciwiwZOUBJMnw7Bhx9zlyy+hS5fwhSSSnU6dciitAPc9PHmy+54WkZii5FhEgiM9
   3U3Ce+EFKF36mLt9+SVccEEY4xLJxjFbumUqU8Z9L/ft6763RSRmqOZYRIJj6FDXtu3TT4+5S2at
   5+bNcNxxYYxN5Cj797sV85YvhypVctixe3c4/XStViMSJVRzLCLhsWIFjBzpRtpy8NVX0KGDEmPx
   3nHHwfnnu7aCORo5EkaNgmXLwhKXiHhPybGIFEzWnsYJCTnuqpIK8ZNOndw/bDlKSHBrTvftC4cO
   hSUuEfGWyipEpGDeeQeef95N/c9hVY/0dNfCLTnZ3c4W8dr69a7r4KZNAe5mHDwI557r/gm84Yaw
   xSciwaeyChEJrbQ017rtlVcCLnc3YwbUr6/EWPyjenU46SSYNSvAjnFx8OqrcO+9rmBeRKKakmMR
   yb+BA+Hqq6Fp04C7fvklXHhhGGISyYOuXeHzz3OxY+PGcM01cOedIY9JRLyl5FhE8ufLL+HHH12X
   ilzurnpj8ZuuXeGLL3K58yOPwPff52IWn4hEMiXHIpJ3O3dCv37w2mtQokTA3VesgD173OCbiJ80
   aQJ//eW+RwMqWdKVEN10kztIRKKSkmMRybt774WOHaFNm1zt/sUXbtTY5DgFQiT8jHHlPrkqrQD3
   fZ+YCIMHhzIsEfGQkmMRyZsZM+Czz+DJJ3N9iEoqxM/yVFoB8Mwz8MknMHNmyGISEe+olZuI5N7f
   f7vVwp58Erp1y9UhO3dCfLxbHa9UqRDHJ5IPe/ZA1arwxx9QrlwuD5owAQYNgoULoXjxkMYnIsGj
   Vm4iElyPPOKKNHOZGIObu9S8uRJj8a8SJaB1a5g0KQ8Hde/ufha0rLRI1FFyLCK58/PP8Oabbjnd
   PPj0U7j44hDFJBIkuW7pltULL8Bbb8HcuaEISUQ8orIKEQls/3446yy46y7o1StPh1WpAosWudIK
   Eb9atw4aNXKr5QVYz+ZI774LTzwB8+ZBkSIhi09EgkNlFSISHE8+CdWquQU/8mDGDKhbV4mx+F98
   PNSqBT/8kMcDr7oKTjgBHn88FGGJiAeUHItIzhYtgmefdf1d89iLTSUVEknyVVphDLz8Mowa5X5W
   RCTiKTkWkWPbvx+uvdbdNq5ZM0+HWqvkWCJLt27uezbPlX7x8fDYY3DDDXDgQEhiE5HwUXIsIsf2
   6KOunOL66/N86Pz5rgtAvXohiEskBBo3drntb7/l4+Abb3R94J54IuhxiUh4KTkWkez9/DO89JJb
   IjofS9tljhprVTyJFMbAJZe49T3ydfAbb8Dzz8MvvwQ9NhEJHyXHIvJv6emunOLZZ6F69XydQiUV
   EonynRwDJCTAU0+5ji7p6UGNS0TCR8mxiPzbww/DKae4mfj5kJwMaWlw9tlBjkskxM4917VzS07O
   5wl69YI6ddzPkIhEJCXHInKk2bPh7bdh9Oh810RMnAgXXQSF9BtGIkyhQu6Ox4QJ+TyBMa6zy9tv
   56MvnIj4gf50icg/9uxx5RSjRkHlyvk+jUoqJJJdckkBkmNwPzsvveR+lnbvDlpcIhIeWiFPRP4x
   cCBs3gzvvZfvU2zeDCef7G5NFy0axNhEwmTfPqha1XWtyGfJvXPNNVC6NLz4YtBiE5GC0Qp5IpJ7
   334LH3/sRo0LYOJE6NhRibFEriJF4MIL3R2QAhk5Ej77DCZPDkpcIhIeSo5FxM2eu+EGeOstKF++
   QKf66CP4z3+CE5aIV7p3L0DXikzHH+/au/XuDdu3ByUuEQk9lVWIxDpr4dJL4aSTXBuqAkhLg9q1
   YcMGtwCISKTas8etf7N6NVSoUMCT3XYbbNwIH3ygxt8iHlNZhYgENmYMrFrlVsMroAkTXEmFEmOJ
   dCVKQLt28PnnQTjZiBGwdKm7MyMivqfkWCSWrVwJ997rJuAFoUhYJRUSTS65xJXhF1ixYvD++3DP
   Pe5nTkR8TWUVIrFq/3447zzo2dPd9i2gzJKK9euhZMkgxCfisZ073aJ3KSlQrlwQTjhqlOt/PGuW
   m/UnImGnsgoRObbhw91f/FtvDcrpMksqlBhLtChTxpVWFKjncVa33AJVqsBDDwXphCISCkqORWLR
   99+7VbzefDNoE4RUUiHR6Ior3Dy6oDDG/cyNHQvTpgXppCISbAGTY2NMJ2PMMmPMSmPMoGy21zPG
   zDbG7DXG3JWXY0XEA2lpcNVVbiJetWpBO+VPP0GXLkE5nYhvXHCB+97esiVIJ6xUySXI114LW7cG
   6aQiEkw5JsfGmDhgFNAJaAD0MMbUP2q3rcCtwFP5OFZEwslauO46NxwWxEx2wgTo0EElFRJ9SpaE
   zp2DNDEvU4cOcPnlcOON7mdSRHwl0MhxMyDZWptird0PjAO6Zd3BWrvFWjsP2J/XY0UkzJ591g3z
   BqFtW1Yffuj+1otEo6CWVmR67DH44w946aUgn1hECipQchwPpGZ5vDbjudwoyLEiEmxz5sB//wvj
   xgV1pvyGDTBvnltuVyQadeoECxe67/WgKVrU/Vf5yCPuB0hEfKNwgO0Fud+T62OHDBly+OPExEQS
   ExMLcFkR+Zc//4Qrr4TRo6FWraCeetw46NYNihcP6mlFfKNYMeja1U06DULXw3/UqeNGji+/HObP
   d8tNi0hQJSUlkZSUlKdjcuxzbIw5Bxhire2U8XgwcMha+0Q2+z4M/GWtfTovx6rPsUiIWev++Fap
   4vqsBlnTpu4OcYcOQT+1iG989ZWrRpo1KwQnv/VWWLfOFTZreWmRkApGn+N5QF1jTC1jTBHgCuCz
   Y12vAMeKSKiMHg3JyfDUU4H3zaPly2HtWmjbNuinFvGVdu1gxQr4/fcQnPyppyA1FZ5/PgQnF5G8
   yjE5ttYeAAYAk4AlwAfW2qXGmL7GmL4AxpiqxphU4A7gAWPMGmNMqWMdG8oXIyJH+fFHGDLE1TYW
   Kxb007/3nqvWKByoQEskwhUp4m7AvPtuCE6eWX/82GPuZ1ZEPKXlo0Wi1ebNruZh1Ci46KKgn95a
   qFsX3n8fzjor6KcX8Z0ff3TtiZctC1H1w6efwu23w4IFUL58CC4gIlo+WiRWHTjghnR79QpJYgyu
   +UWhQi7/FokFZ5/t/imcOzdEF7j4YrjsMujZEw4eDNFFRCQQJcci0eiBByAuDoYODdkl3nvPLbSn
   +UMSK4xx/2++804IL/Lf/8LevfDwwyG8iIjkRGUVItFmwgQYOBB+/hkqVgzJJQ4cgBo1YOZMV1oh
   Eit+/x2aNXPNJYLYLvxImze7WqXnnoPu3UN0EZHYpLIKkVizYgX07esasoYoMQb45hs48UQlxhJ7
   TjwR6tVzPwMhU7mya+vWty8s1Tx2kXBTciwSLXbudKNMw4a5oa0QevNNuP76kF5CxLeuuQbGjg3x
   RZo2hREjXB3yjh0hvpiIZKWyCpFocPCgS4yrVYOXXw5pIfCWLW7E+I8/oGzZkF1GxLf+/NMtNPn7
   71CuXIgvNmAArFnjOlkU0niWSEGprEIkVjzwgBtdeuGFkM+Qe/dduPBCJcYSu44/3q0I+cEHYbjY
   M8/Atm0hnVwrIkdSciwS6d57z/2V/vjjEM4QcqxVSYUIwA03wBtvhOFCRYrA+PEwZoybSyAiIafk
   WCSSzZ3rOlNMnBjSCXiZFixwA9Rt2oT8UiK+1r49bNoECxeG4WJVq7qf8f79Q9hkWUQyKTkWiVTr
   17s649deg4YNw3LJN990K4Sp9FFiXVxcGEePAZo0cT/r3bvD2rVhuqhIbNKEPJFI9Pff0Lq1m8l+
   331huWR6OsTHu4GrE08MyyVFfO2PP+DMMyE1FYoXD9NFR4xwa7bPnAmlSoXpoiLRQxPyRKLRoUNu
   ma66dWHw4LBdduJEaNRIibFIphNOcB3XJkwI40XvvtuNIvfq5X4XiEjQKTkWiTR33w1paW6CThjX
   bn7lFbjpprBdTiQi3HgjvP56GC9ojGvXuG1b2O4aicQalVWIRJIXXoDRo2HWrDA0WP3H8uXQqpVr
   t1q0aNguK+J7+/ZBQoL7kaxTJ4wXTkuDc86BQYOgT58wXlgksqmsQiSaTJwI//0vfPVVWBNjcKPG
   N9ygxFjkaEWKuAqHsE3My1SxInz9NTz0EHz+eZgvLhLdNHIsEgl++smtvPH1167IMYz+/tuNjGki
   nkj2li9382P/+MODfyDnzHG/Gz77zI0ki0iONHIsEg1WrXJdKd58M+yJMcCHH0KzZkqMRY7llFNc
   N8Xx4z24eLNm8NZb7nfE8uUeBCASfZQci/jZxo3QsaO7dXrhhZ6E8PLLcPPNnlxaJGIMGACjRnl0
   8S5d4PHHoVMn2LDBoyBEooeSYxG/2r7dJcbXXgv9+nkSwi+/wLp1cMEFnlxeJGJceKFbl+fnnz0K
   4PrroXdvlyjv3OlRECLRQcmxiB/t3u3+2rZtCw884FkYL7/sJsLHxXkWgkhEiItz/8O++KKHQdx/
   v6s77tbNTRYQkXzRhDwRv9m3Dy66CKpWdb2MPVqreds215pqyRIXiojkbMsWtzbPqlVQoYJHQRw8
   6Npn7NjhVicpUsSjQET8SRPyRCJN5h+2YsXcygIeJcYAr776T44uIoFVquQGbceM8TCIuDh4+204
   7jjo2RMOHPAwGJHIpJFjEb+w1s18W7nS9TIuVsyzUPbvd90pvvgCGjf2LAyRiDN3Llx+OSQne1yO
   lJ7u/rutVs3TO1AifqORY5FIYS3cdhssXOgW+/AwMQb46CN3e1iJsUjenHWWu9vy6aceB1K0qCur
   WLUKbr3V/Y4RkVxRciziNWvhjjvcQh+TJkHp0p6H8+yzLiQRybv/+z948kkf5KMlSrjbP3PmuGWm
   PQ9IJDIoORbxkrXuL+n338O330LZsl5HxA8/uC5yHrVVFol4F18MaWkwa5bXkeB+p3zzjfv9cu+9
   SpBFckHJsYhXrHWjOdOnuz9cxx/vdUSAGzW+/XaVKIrkV1wc3HknPPWU15FkqFABpk6FyZPdP+NK
   kEVypAl5Il6w1o3iTJrk/mh51vfpSKtWwdlnQ0oKlCrldTQikWvPHqhVC2bOdMtL+8L27dChAzRv
   Ds89BybHOUkiUUkT8kT86NAhNzQ7daqvEmOAESNcwwwlxiIFU6KEWxTkmWe8jiSLcuXc6PGPP8It
   t7jfRSLyLxo5FgmngwfdknPLl7t2bT6oMc60bh00bAgrVkDFil5HIxL5Nm92o8bLl0Plyl5Hk8WO
   HdC5s/uBHz1aNVQSUzRyLOIn+/e7pvxr1vhm8l1WTz8N116rxFgkWCpXhiuvdBUMvlK2rCvpWroU
   rn4G3JIAABl/SURBVLnG/W4SkcM0ciwSDnv3whVXuJHj8eM972N8tLQ0OPlkWLQI4uO9jkYkeqSk
   wJlnurV9ypf3Opqj7NnjViyx1jU3L1HC64hEQk4jxyJ+kHkLs1gx+OQT3yXGAM8/D5ddpsRYJNhq
   1YJLLnFdYHynRAm3UEiFCtC+PWzb5nVEIr6gkWORUFq/3iXGrVu7v46eriebvZ074aST3BoktWt7
   HY1I9Fm92q2cl5zs5sT5zqFDrsXb5Mmu3KJ6da8jEgkZjRyLeGnpUtcy6aqr3NCsDxNjgFGjoGNH
   JcYioXLSSdCtm/s14EuFCrlJB1ddBeed52pARGKYRo5FQmH2bOjeHZ54ws1y86nt212t8axZ7r2I
   hEZmD/HkZN+s95O9116DBx+Ejz+GFi28jkYk6DRyLOKFjz5yw0RvvunrxBjcCl4XXaTEWCTUateG
   Cy6AkSO9jiSAPn3grbfcP/fvv+91NCKe0MixSLBYC8OHw+uvw8SJ0Lix1xHlaNMmaNAA5s+HE07w
   OhqR6Ldypau0WrbMV2v/ZG/RIujaFXr3hgce0Gp6EjVyM3Ks5FgkGPbudX9EkpNdYly1qtcRBTRw
   oJuH4/uRLJEo0r8/FC/uSnx9b+NGd2upXj1XblG0qNcRiRRYUJJjY0wn4DkgDnjdWvtENvuMBDoD
   e4DrrLULMp5PAXYCB4H91tpm2Ryr5Fgi28aNcPHFcOKJMGaM+8vnc2vWQJMmsHhxROTxIlFj40Y4
   9dQIumOzZ49bKGTLFlcy5qul/kTyrsA1x8aYOGAU0AloAPQwxtQ/ap8uQB1rbV3gJmB0ls0WSLTW
   NskuMRaJeL/8Auec49q1vfdeRCTGAMOGwU03KTEWCbeqVeGWW9yct4hQogR8+CG0bOn60c2b53VE
   IiEXaEJeMyDZWptird0PjAO6HbXPRcDbANban4DjjTFVsmxXoZJEp7ffdo3zR4yAhx+OmJq8X3+F
   zz6De+7xOhKR2PR//+dWkF+40OtIcqlQITef4rnnoEsXN9lYJIoFSo7jgdQsj9dmPJfbfSwwxRgz
   zxjTpyCBivhGejr06wePPQZJSW751QhhLdx5Jzz0kE8XIxCJAWXKwP33w+DBXkeSR927w3ffuRaV
   /fvDvn1eRyQSEoGS49wWAx9ryOw8a20TXD3yLcaYlrmOTMSP1qyBVq1cq4e5c13xYAT5/HPYsAH6
   9vU6EpHY1rcvrFjhFqSLKPXru+U0162DNm3cKqAiUaZwgO3rgIQsjxNwI8M57VMj4zmstesz3m8x
   xkzAlWnMPPoiQ4YMOfxxYmIiiYmJuQpeJKymTIGrr4a77nL3RSOkjCLTvn0u9FGjoHCgn3wRCaki
   RVyVwm23ua5pRYp4HVEelC0LEyb8f3t3Hh5ldbYB/H5kESWA7IJBQQsKKJvsQgFBBaQqqBUKVlHU
   2oJLvyouVz/9rC18RQErVq1QRUSpAu6AoBK3imwBDCAEKiigAQQTEoiE5PSPe8ZJYlaynHkn9++6
   zvW+M5mBBybzzjNneQ5Hz7p04TSLIUN8RyVSoISEBCQkJJTqOUVWqzCz6gC2ABgIYA+AlQBGOec2
   53rMUADjnXNDzawngOnOuZ5mdjKAas65Q2ZWG8BSAP/nnFua7+9QtQqJbllZnIfw/PPAnDnAhRf6
   jui4TJ0KvPce8PbbviMRkbBLLwX69wfuust3JMfpgw/YaTByJPDnPwcsy5eqqLxKuQ1BpJTbLOfc
   JDO7BQCcc0+HHhOuaJEBYKxzbq2ZnQlgYeiPqQ5grnNuUgF/vpJjiV7btwOjRgGNG7N3JKBljPbu
   5QyQjz5iyVIRiQ7JyUCvXlycd1r+FT1BsX8/cP31LPf20kvAmWf6jkikUNoERKQs5szh6rU//hGY
   MCFw0yhyu/ZalpCaMsV3JCKS3/33Azt2AHPn+o6kDJwDpk0DJk3i3K1rrvEdkUiBlByLHI/vvwfG
   j2eV/pdeAjp29B1Rmbz7LjBuHDf8qF3bdzQikl9GBte5zZkD9OvnO5oyWrWK0yzOP59JcoMGviMS
   yaPMm4CIVDmLFwPnncdaS6tXBz4xzsxkxaUZM5QYi0Sr2rWBxx7jxjxHjviOpoy6dQMSEzkVrUMH
   YNEi3xGJlJp6jkUAIDWVUyjefx+YNSuwi+7ye+ABICkJWLDAdyQiUpyrrwbOOguYPNl3JOVk+XJg
   7Fjg4ouBRx8F6tTxHZGIeo5FSuSdd9hbXKMGt4+LkcR4/XrgySfZIyUi0W/GDK77XbPGdyTlZMAA
   XlNzctiLvHRp8c8RiQLqOZaqa98+1k9avpy9xYMG+Y6o3Bw9CnTvDtxxBxeRi0gwzJkDPPIIp+7G
   VFW0JUu4s+gFF7CuZEAr/0jwqedYpCA5OUyGzz2XeygnJcVUYgyw3GiLFsB11/mORERKY8wYID4+
   hqZWhA0ezGtts2YcqXv2WVa4EIlC6jmWqiUpib0XR48CTz0FdO7sO6Jyt2YNMHQo18Q0b+47GhEp
   rV27uPHcW29xBCjmrFsH3HQTEBfHuV8qvi6VSD3HImFpacDEiZxPPHo08O9/x2RinJ4O/OpX3JZW
   ibFIMMXHA3//O/cfSkvzHU0F6NQJWLECuOIKoG9f7mufmuo7KpEfKTmW2JadDcycCZx9NpCSwsUh
   v/kNUK2a78gqxIQJQO/e/FAVkeC66ip+l//d73xHUkGqVQNuv50F2NPS2Hs8cyav2SKeaVqFxK7l
   y4E772T5oOnTWZQ+hr34IvDQQyzPHBfnOxoRKauMDKBrV+6gN2aM72gq2Jo1TJaPHGGJnT59fEck
   MUo75EnV9MUXwL33cl7blCnAlVcGeuvnkti6lYvAly3jiKWIxIZ164CLLgI++QRo08Z3NBXMOWDe
   PE6B69wZ+MtfgPbtfUclMUZzjqVq2bkTuOEGzmHr2RPYvJljkzGeGB86BAwfDjz8sBJjkVjTqROr
   z1xxRYzOP87NjHPCtm4Ffv5z1kkeOxb46ivfkUkVo+RYgi8lhcNxXbpwFVpyMnseatXyHVmFc451
   jHv3Bm65xXc0IlIRbr6ZueKvf81KlDGvVi0u0ktO5jW9c2fe3rfPd2RSRSg5luBKSQHuuQdo1449
   Dps2sfv0lFN8R1ZpJk8Gdu/mzloiErv+9jdg/37gT3/yHUklqleP3eZJSUBmJhdW/+EPvPaLVCAl
   xxI8X38N3HYb0LYt5xQkJnLBXdOmviOrVPPnA088weOJJ/qORkQqUs2afK/PnAm8+qrvaCpZs2a8
   2G3YAPzwA6/9d94J7NnjOzKJUUqOJTi2bQPGjeMkvFq1WALoiSeA00/3HVml+/RT7mXy5pusiSoi
   se/UU5kY33wzS7VXOfHxwOOPsycZ4C6nv/0tp1+IlCMlxxL9VqwArrkG6NWLF8fkZOCvf2VvQhW0
   bRswYgQwe3ZM7mMiIkXo2hWYM4fXgM2bfUfjSfPmwLRpnErXoAEXXQwfzpIeqn4l5UCl3CQ6ZWVx
   DHH6dOC777i7xQ03sGZxFbZ7N4tx3HMPe49EpGqaPRt44AHmg6ed5jsazzIygOeeY8LcqBEX7w0f
   DlSv7jsyiUKqcyzBs38/8I9/cO/UNm2AO+4ALr00Zne0K429e4F+/VjZ6O67fUcjIr5Nngy88AL3
   O2rc2Hc0USA7G3jjDeDRR7k25eabgRtv5HwUkRDVOZZgyMkB3n+f9S1/9jPOG1i0iPdddpkSYwAH
   DgCXXAJcfbUSYxGhiRNZ//jCC/nlucqrVo09xh9/DLz2Gmvft20L/PKX/AahjjgpIfUciz8pKRwK
   mzkTOOkkfssfPRqoX993ZFFl717ukHXxxZxqHeN7mohIKTjH6RULF7I/oUkT3xFFmdRUTtJ+8kn2
   LIc/Z6pYdSOJ0LQKiT6ZmcDbb0fGAq+8ErjpJqBHD2V9Bdi9Gxg0iB0fDz6o/yIRKdiDDwKvvAK8
   +26VXatcNOeAjz4CZs0CXn8d6NMHuO464Be/qBIbRkmEkmOJDjk5wIcfMiFeuJCl2EaP5tbO9er5
   ji5qbdkCDB3Kjo6JE31HIyLRbtIk4OmngcWLOZtACpGeDixYwFWN69ez9+Haa4GePYETNNs01ik5
   Fn+c4+Yc//oX8OKLQMOGTIhHjVJh3hL44ANWr5s0iQvwRERK4vnngbvuYrGfvn19RxMAO3ey42bu
   XCbNV13FZFmjmTFLybFUruxs7k6xcCFbjRq80IwezWLtUiJz5rAS0UsvAQMH+o5GRIJm2TJedh97
   jP0RUkIbNwIvv8x2+HAkUe7eXYlyDFFyLBXvhx+AhARu2/TaayyZM2IEW/v2uqCUQlYWcN997PF5
   +22gXTvfEYlIUG3YwMINl13Ghbw1aviOKECciyTKr7wCpKUBw4axDRwInHyy7wilDJQcS8X48ktO
   alu8mOP/553Hq/Dw4cBZZ/mOLpB27QJGjgTq1mXPccOGviMSkaA7eBAYM4a53csva6Hecdu6FXjz
   TeCtt4A1a1hwPpwsV/kdWIJHybGUj4wMbsO0ZAkT4gMHgMGDgSFDWF+sQQPfEQbakiWcV3zbbVx4
   p/UgIlJecnKAhx8GnnqK+ysNG+Y7ooA7eBB45x0my0uW8BvHoEFs/fpV+V1cg0DJsRyfzEzOHV6+
   nC0xEejcmYnwkCFAly7K4MpBWhrnFi9dykXT/fv7jkhEYtUHHwDXX88cbupU5XDlIjsbWLuW9fOW
   LQNWruRnZThZ7t5d81mikJJjKZn0dL6pP/mEyfDKlVxAN2AAt17q3RuoXdt3lDFl2TJg3DjuevfI
   I5xOISJSkdLSgN//npuFPP00NxeScnT4MHfnCyfL27YB3bqxpnKfPkCvXvpWEgWUHMtPOQds386e
   4XDbupW1h3v3Zvdl377K1irIV1+xzNKKFRzivOQS3xGJSFWzaBEwYQIHAadOBVq08B1RjPr+e37G
   fvQRk+a1a4Gzz2aifMEFTJxbttTC9Uqm5FiAb77hG3LtWmD1ar5RTzyR32B79WJC3KkT75MKc/gw
   8OijLK00fjxw991a8Cwi/hw5wioWjz8O3HEHW1yc76hi3A8/cEHfxx9zpHbVKpYp6tqViXK3bjzX
   yskKpeS4KnEO2LGD84PDyfDatcCxY8D557OLoEsX7gCkboJKk5nJHuLJk9lZMGUKcMYZvqMSEaHt
   24H77+ec5Pvu446c6iupRHv2MEletYodWKtWASedxCS5Y0dWg+rQgZWgqlXzHW1MUHIci5zjm2nj
   RiApicdwq1s3kgSHW3y8hmw8yMgAnnuOSXGnTsBDD3GdhohINEpMZJK8aRPnJd94o5aaeOEcy6Wu
   WQN8/jkLVm/YAKSkcO+ADh0irX17oFEjfcaXkpLjIMvMBP7zHyA5mW3rVl61Nm4Eatbkm6J9ey6c
   C5/Xr+876ipvzx5gxgzgmWc4dXviRO5CKiISBCtWcITrww+BW24Bbr1VpXyjQloaO8Q2bGDSvH49
   cwKA85hzt3POYU+zhgAKpOQ42qWnc1/3L7+MJMHh9u23HH9v3TrS2rVjEty4se/IJZesLC5wefZZ
   Dk1eey1w++3aD0VEgis5GZg2DZg3j1PCxo0Dhg4Fqlf3HZn8yDlg/35gyxbgiy94DJ9/9RVHjtu0
   Ac48E2jVii18Xq+e7+i9UXLsk3PcLGPnzp+2HTt4PHKECXDLlnmT4Nateb+uQlErJwf47DNu9Tx3
   Ll+ysWOBq69WpR4RiR0ZGdxB+Zln2I8zciRw5ZVcz61y91Hs6FGOPm/dyuOXX+ZtNWtGEuZwa9GC
   wwTx8ZyuEaMvsJLjinLoEKtA7NkTOeY/37OHxb/POKPg1rKl5goFTGYmhxpffRV4/XVuDDhiBHuK
   W7f2HZ2ISMXavJnbUM+fz76fESO4417fvqq+EyjhHuf8CfPu3cCuXWzp6ayaER8fSZjD582bA02b
   ssXFBS6PUXJcUpmZwL59/GXZt++n5+HbKSlMerOz+QvSrBl/SZo3L/hctYID7ehRYN064L33WNN9
   5UouHL7iCmD4cCXEIlJ1bdkCLFgALF7M62SPHtxEtV8/VQeNCUeOMN/ZtStv0rx7N9vevcyJnGOS
   3KRJJGHOfd60KaeCNmzIHqWaNX3/y8onOTazwQCmA6gGYKZz7v8LeMzfAAwBcBjA9c65xFI8t3yS
   46wsIDWVRbcPHsx7LOi+AwciCfDRo+zFbdyYrbDzJk0iSW/AvilJ0bKz+cV59WpOl/jsM653OOss
   bhI4cCAv+vq+IyKSV1oakJDATeE+/pgj+R07cupFjx5MllWJLEalp0cS5ZSUvOfh23v3Mvc6cIDJ
   cYMGbOGEubB2yin80A23ctqKu8zJsZlVA7AFwCAAuwGsAjDKObc512OGAhjvnBtqZj0APOac61mS
   54ae79yWLfwPPnSo8GNhP0tLY7J75Aj/I8Otfv28x4LuCye9deoo2T1OCQkJ6N+/v+8wSiwjg9O9
   t2/PWwVvyxZ+D+rShRfzHj1YZjKW5w8H7bWTvPT6BVssv37p6SzX++mn7Gj4/HPmSeecwwJL557L
   8/BU16CVjIvl165COcdfjgMHCm7ffZf3dmoqc7xwq149b7Jcty4XFua/r25dTveoXZvHfOfWrFmx
   yXFxK766A9jmnNsBAGY2D8DlAHInuJcBmM1/t/vMzE4xs1MBtCrBc+nSS5mFxMUVfGzUiHN0C/pZ
   nTpMdgM47yUWRMtFIjub76vcX1hTUjgNPLwGcscOfqc64wxekNu1AwYM4I51bdtWvV7haHnt5Pjo
   9Qu2WH794uJ4bR0wIHLfoUOsPJaUxGQ5IYHrxHbsYH7TqhU/5ps1i7RTT42c168fPR/xsfzaVSiz
   SN5W2t2wnOMU2LS0nybN4Zaayh7qnTvZE5aezhY+Dx9LoLjk+DQAX+e6vQtA/qqtBT3mNADNS/Bc
   Sk4uQagSZM4xgc3O5qZ9WVns7M/MLP546BB/53O38Psg3A4c4AU29zSnJk14UT3/fF50W7bkfTG6
   AFdEJGrVqRMZlcstJ4eVS8OJ8rffslMjMTFy/s03zGnq1WOSHG65B4Jr1+aiwJNPLvz8xBM5Ml+j
   Bkf3c59Xrx49ybcUwIw7B550Ej/gy/pnFaO45Likk4HL9Cs1bFjoL8v3t0XT7WiKJRpiPXaMiW5K
   CjB7diTpDd8fPg/fzs5mUlq9eqSFf89r1Yocc5+Hj3FxvCiefjqPBbVGjVT5TkQkaE44IbKWvU+f
   wh8XXlZ08GDeJUQHD/L+w4eZTB8+zJaRkfc8I4PLi7KyIsfc58eO8TMkf+J8wgk/bQcPsv5zQT8r
   quXPyUpzO1qeW1UUN+e4J4AHnXODQ7fvBZCTe2GdmT0FIME5Ny90+wsA/cBpFUU+N3R/wOq4iYiI
   iEhQlXXO8WoArc2sJYA9AK4BMCrfY94AMB7AvFAy/b1zLsXMvivBc4sNUERERESkshSZHDvnjpnZ
   eADvgOXYZjnnNpvZLaGfP+2cW2RmQ81sG4AMAGOLem5F/mNERERERMrC+yYgIiIiIiLRImrW7ZvZ
   BDPbbGZJZvaTzUIkupnZ/5hZjpk18B2LlJyZTQm979ab2UIzq+c7JimamQ02sy/MLNnMJvqOR0rO
   zFqY2XIz2xj6rLvNd0xSemZWzcwSzexN37FIyYVKDc8PfeZtCk0FLlBUJMdmNgCsl9zBOXcugEc8
   hySlYGYtAFwEYKfvWKTUlgJo75zrCGArgHs9xyNFCG2uNAPAYADtAIwys7Z+o5JSyAJwp3OuPYCe
   AH6n1y+QbgewCSWv6CXR4TEAi5xzbQF0QEH7boRERXIM4FYAk5xzWQDgnNvnOR4pnakA7vYdhJSe
   c26Zcy4ndPMzAPE+45Fi/bgxU+h6Gd5cSQLAOfetc25d6Dwd/HBu7jcqKQ0ziwcwFMBMlLGMrVSe
   0KhoX+fcPwGui3POpRb2+GhJjlsD+LmZrTCzBDPr6jsgKRkzuxzALufcBt+xSJndAGCR7yCkSIVt
   uiQBE6rk1Bn8UirBMQ3AXQByinugRJVWAPaZ2bNmttbMnjGzkwt7cKVtm2BmywCcWsCP7g/FUd85
   19PMugF4GcCZlRWbFK2Y1+5eABfnfnilBCUlVsTrd59z7s3QY+4HcNQ592KlBielpWHcGGBmcQDm
   A7g91IMsAWBmwwDsdc4lmll/3/FIqVQH0AXAeOfcKjObDuAeAP9b2IMrhXPuosJ+Zma3AlgYetyq
   0MKuhs657yorPilcYa+dmZ0Lfhtbb9xGJx7AGjPr7pzbW4khShGKeu8BgJldDw4TDqyUgKQsdgNo
   ket2C7D3WALCzGoAWADgBefca77jkVLpDeAyMxsKoBaAumb2vHPu157jkuLtAke5V4VuzweT4wJF
   y7SK1wBcCABm1gZATSXG0c85l+Sca+qca+WcawX+8nVRYhwcZjYYHCK83DmX6TseKdaPGzOZWU1w
   c6U3PMckJWTsRZgFYJNzbrrveKR0nHP3OedahD7vRgJ4X4lxMDjnvgXwdSjHBIBBADYW9vhK6zku
   xj8B/NPMPgdwFIB+2YJJQ77B8ziAmgCWhXr/P3XO/dZvSFIYba4UeBcAGANgg5klhu671zm3xGNM
   cvz0mRcsEwDMDXUsbEdo07qCaBMQEREREZGQaJlWISIiIiLinZJjEREREZEQJcciIiIiIiFKjkVE
   REREQpQci4iIiIiEKDkWEREREQlRciwiIiIiEqLkWEREREQk5L+47bX/+WXVtwAAAABJRU5ErkJg
   gg==
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>Compare the estimates of the Logit Fair model above to a Probit model. Does the prediction table look better? Much difference in marginal effects?</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h3 id="Genarlized-Linear-Model-Example">Genarlized Linear Model Example<a class="anchor-link" href="#Genarlized-Linear-Model-Example">&#182;</a></h3>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[19]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">star98</span><span class="o">.</span><span class="n">SOURCE</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   
   Jeff Gill&apos;s &#96;Generalized Linear Models: A Unified Approach&#96;
   
   http://jgill.wustl.edu/research/books.html
   
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[20]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">star98</span><span class="o">.</span><span class="n">DESCRLONG</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   
   This data is on the California education policy and outcomes (STAR program
   results for 1998.  The data measured standardized testing by the California
   Department of Education that required evaluation of 2nd - 11th grade students
   by the the Stanford 9 test on a variety of subjects.  This dataset is at
   the level of the unified school district and consists of 303 cases.  The
   binary response variable represents the number of 9th graders scoring
   over the national median value on the mathematics exam.
   
   The data used in this example is only a subset of the original source.
   
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[21]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">star98</span><span class="o">.</span><span class="n">NOTE</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   ::
   
       Number of Observations - 303 (counties in California).
   
       Number of Variables - 13 and 8 interaction terms.
   
       Definition of variables names::
   
           NABOVE   - Total number of students above the national median for the
                      math section.
           NBELOW   - Total number of students below the national median for the
                      math section.
           LOWINC   - Percentage of low income students
           PERASIAN - Percentage of Asian student
           PERBLACK - Percentage of black students
           PERHISP  - Percentage of Hispanic students
           PERMINTE - Percentage of minority teachers
           AVYRSEXP - Sum of teachers&apos; years in educational service divided by the
                   number of teachers.
           AVSALK   - Total salary budget including benefits divided by the number
                      of full-time teachers (in thousands)
           PERSPENK - Per-pupil spending (in thousands)
           PTRATIO  - Pupil-teacher ratio.
           PCTAF    - Percentage of students taking UC/CSU prep courses
           PCTCHRT  - Percentage of charter schools
           PCTYRRND - Percentage of year-round schools
   
           The below variables are interaction terms of the variables defined
           above.
   
           PERMINTE_AVYRSEXP
           PEMINTE_AVSAL
           AVYRSEXP_AVSAL
           PERSPEN_PTRATIO
           PERSPEN_PCTAF
           PTRATIO_PCTAF
           PERMINTE_AVTRSEXP_AVSAL
           PERSPEN_PTRATIO_PCTAF
   
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[22]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">dta</span> <span class="o">=</span> <span class="n">sm</span><span class="o">.</span><span class="n">datasets</span><span class="o">.</span><span class="n">star98</span><span class="o">.</span><span class="n">load_pandas</span><span class="p">()</span><span class="o">.</span><span class="n">data</span>
   <span class="k">print</span><span class="p">(</span><span class="n">dta</span><span class="o">.</span><span class="n">columns</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   Index([u&apos;NABOVE&apos;, u&apos;NBELOW&apos;, u&apos;LOWINC&apos;, u&apos;PERASIAN&apos;, u&apos;PERBLACK&apos;, u&apos;PERHISP&apos;, u&apos;PERMINTE&apos;, u&apos;AVYRSEXP&apos;, u&apos;AVSALK&apos;, u&apos;PERSPENK&apos;, u&apos;PTRATIO&apos;, u&apos;PCTAF&apos;, u&apos;PCTCHRT&apos;, u&apos;PCTYRRND&apos;, u&apos;PERMINTE_AVYRSEXP&apos;, u&apos;PERMINTE_AVSAL&apos;, u&apos;AVYRSEXP_AVSAL&apos;, u&apos;PERSPEN_PTRATIO&apos;, u&apos;PERSPEN_PCTAF&apos;, u&apos;PTRATIO_PCTAF&apos;, u&apos;PERMINTE_AVYRSEXP_AVSAL&apos;, u&apos;PERSPEN_PTRATIO_PCTAF&apos;], dtype=&apos;object&apos;)
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[23]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">dta</span><span class="p">[[</span><span class="s">&#39;NABOVE&#39;</span><span class="p">,</span> <span class="s">&#39;NBELOW&#39;</span><span class="p">,</span> <span class="s">&#39;LOWINC&#39;</span><span class="p">,</span> <span class="s">&#39;PERASIAN&#39;</span><span class="p">,</span> <span class="s">&#39;PERBLACK&#39;</span><span class="p">,</span> <span class="s">&#39;PERHISP&#39;</span><span class="p">,</span> <span class="s">&#39;PERMINTE&#39;</span><span class="p">]]</span><span class="o">.</span><span class="n">head</span><span class="p">(</span><span class="mi">10</span><span class="p">))</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
      NABOVE  NBELOW    LOWINC   PERASIAN   PERBLACK    PERHISP   PERMINTE
   0     452     355  34.39730  23.299300  14.235280  11.411120  15.918370
   1     144      40  17.36507  29.328380   8.234897   9.314884  13.636360
   2     337     234  32.64324   9.226386  42.406310  13.543720  28.834360
   3     395     178  11.90953  13.883090   3.796973  11.443110  11.111110
   4       8      57  36.88889  12.187500  76.875000   7.604167  43.589740
   5    1348     899  20.93149  28.023510   4.643221  13.808160  15.378490
   6     477     887  53.26898   8.447858  19.374830  37.905330  25.525530
   7     565     347  15.19009   3.665781   2.649680  13.092070   6.203008
   8     205     320  28.21582  10.430420   6.786374  32.334300  13.461540
   9     469     598  32.77897  17.178310  12.484930  28.323290  27.259890
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[24]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">dta</span><span class="p">[[</span><span class="s">&#39;AVYRSEXP&#39;</span><span class="p">,</span> <span class="s">&#39;AVSALK&#39;</span><span class="p">,</span> <span class="s">&#39;PERSPENK&#39;</span><span class="p">,</span> <span class="s">&#39;PTRATIO&#39;</span><span class="p">,</span> <span class="s">&#39;PCTAF&#39;</span><span class="p">,</span> <span class="s">&#39;PCTCHRT&#39;</span><span class="p">,</span> <span class="s">&#39;PCTYRRND&#39;</span><span class="p">]]</span><span class="o">.</span><span class="n">head</span><span class="p">(</span><span class="mi">10</span><span class="p">))</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
      AVYRSEXP    AVSALK  PERSPENK   PTRATIO     PCTAF  PCTCHRT   PCTYRRND
   0  14.70646  59.15732  4.445207  21.71025  57.03276        0  22.222220
   1  16.08324  59.50397  5.267598  20.44278  64.62264        0   0.000000
   2  14.59559  60.56992  5.482922  18.95419  53.94191        0   0.000000
   3  14.38939  58.33411  4.165093  21.63539  49.06103        0   7.142857
   4  13.90568  63.15364  4.324902  18.77984  52.38095        0   0.000000
   5  14.97755  66.97055  3.916104  24.51914  44.91578        0   2.380952
   6  14.67829  57.62195  4.270903  22.21278  32.28916        0  12.121210
   7  13.66197  63.44740  4.309734  24.59026  30.45267        0   0.000000
   8  16.41760  57.84564  4.527603  21.74138  22.64574        0   0.000000
   9  12.51864  57.80141  4.648917  20.26010  26.07099        0   0.000000
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[25]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">formula</span> <span class="o">=</span> <span class="s">&#39;NABOVE + NBELOW ~ LOWINC + PERASIAN + PERBLACK + PERHISP + PCTCHRT &#39;</span>
   <span class="n">formula</span> <span class="o">+=</span> <span class="s">&#39;+ PCTYRRND + PERMINTE*AVYRSEXP*AVSALK + PERSPENK*PTRATIO*PCTAF&#39;</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h4 id="Aside:-Binomial-distribution">Aside: Binomial distribution<a class="anchor-link" href="#Aside:-Binomial-distribution">&#182;</a></h4>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>Toss a six-sided die 5 times, what&#39;s the probability of exactly 2 fours?</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[26]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">stats</span><span class="o">.</span><span class="n">binom</span><span class="p">(</span><span class="mi">5</span><span class="p">,</span> <span class="mf">1.</span><span class="o">/</span><span class="mi">6</span><span class="p">)</span><span class="o">.</span><span class="n">pmf</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[26]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0.16075102880658435
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[27]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="kn">from</span> <span class="nn">scipy.misc</span> <span class="kn">import</span> <span class="n">comb</span>
   <span class="n">comb</span><span class="p">(</span><span class="mi">5</span><span class="p">,</span><span class="mi">2</span><span class="p">)</span> <span class="o">*</span> <span class="p">(</span><span class="mi">1</span><span class="o">/</span><span class="mf">6.</span><span class="p">)</span><span class="o">**</span><span class="mi">2</span> <span class="o">*</span> <span class="p">(</span><span class="mi">5</span><span class="o">/</span><span class="mf">6.</span><span class="p">)</span><span class="o">**</span><span class="mi">3</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[27]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0.1607510288065844
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[28]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="kn">from</span> <span class="nn">statsmodels.formula.api</span> <span class="kn">import</span> <span class="n">glm</span>
   <span class="n">glm_mod</span> <span class="o">=</span> <span class="n">glm</span><span class="p">(</span><span class="n">formula</span><span class="p">,</span> <span class="n">dta</span><span class="p">,</span> <span class="n">family</span><span class="o">=</span><span class="n">sm</span><span class="o">.</span><span class="n">families</span><span class="o">.</span><span class="n">Binomial</span><span class="p">())</span><span class="o">.</span><span class="n">fit</span><span class="p">()</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[29]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="n">glm_mod</span><span class="o">.</span><span class="n">summary</span><span class="p">())</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
                     Generalized Linear Model Regression Results                   
   ================================================================================
   Dep. Variable:     [&apos;NABOVE&apos;, &apos;NBELOW&apos;]   No. Observations:                  303
   Model:                              GLM   Df Residuals:                      282
   Model Family:                  Binomial   Df Model:                           20
   Link Function:                    logit   Scale:                             1.0
   Method:                            IRLS   Log-Likelihood:                -2998.6
   Date:                  Thu, 21 May 2015   Deviance:                       4078.8
   Time:                          05:57:21   Pearson chi2:                 4.05e+03
   No. Iterations:                       7                                         
   ============================================================================================
                                  coef    std err          z      P&gt;|z|      [95.0% Conf. Int.]
   --------------------------------------------------------------------------------------------
   Intercept                    2.9589      1.547      1.913      0.056        -0.073     5.990
   LOWINC                      -0.0168      0.000    -38.749      0.000        -0.018    -0.016
   PERASIAN                     0.0099      0.001     16.505      0.000         0.009     0.011
   PERBLACK                    -0.0187      0.001    -25.182      0.000        -0.020    -0.017
   PERHISP                     -0.0142      0.000    -32.818      0.000        -0.015    -0.013
   PCTCHRT                      0.0049      0.001      3.921      0.000         0.002     0.007
   PCTYRRND                    -0.0036      0.000    -15.878      0.000        -0.004    -0.003
   PERMINTE                     0.2545      0.030      8.498      0.000         0.196     0.313
   AVYRSEXP                     0.2407      0.057      4.212      0.000         0.129     0.353
   PERMINTE:AVYRSEXP           -0.0141      0.002     -7.391      0.000        -0.018    -0.010
   AVSALK                       0.0804      0.014      5.775      0.000         0.053     0.108
   PERMINTE:AVSALK             -0.0040      0.000     -8.450      0.000        -0.005    -0.003
   AVYRSEXP:AVSALK             -0.0039      0.001     -4.059      0.000        -0.006    -0.002
   PERMINTE:AVYRSEXP:AVSALK     0.0002   2.99e-05      7.428      0.000         0.000     0.000
   PERSPENK                    -1.9522      0.317     -6.162      0.000        -2.573    -1.331
   PTRATIO                     -0.3341      0.061     -5.453      0.000        -0.454    -0.214
   PERSPENK:PTRATIO             0.0917      0.015      6.321      0.000         0.063     0.120
   PCTAF                       -0.1690      0.033     -5.169      0.000        -0.233    -0.105
   PERSPENK:PCTAF               0.0490      0.007      6.574      0.000         0.034     0.064
   PTRATIO:PCTAF                0.0080      0.001      5.362      0.000         0.005     0.011
   PERSPENK:PTRATIO:PCTAF      -0.0022      0.000     -6.445      0.000        -0.003    -0.002
   ============================================================================================
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>The number of trials </p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[30]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">glm_mod</span><span class="o">.</span><span class="n">model</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">orig_endog</span><span class="o">.</span><span class="n">sum</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[30]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0      807
   1      184
   2      571
   3      573
   4       65
   5     2247
   6     1364
   7      912
   8      525
   9     1067
   10    3016
   11     235
   12     556
   13     688
   14     252
   ...
   288      53
   289     266
   290     304
   291    1338
   292    1170
   293    1431
   294     248
   295     516
   296     591
   297      59
   298     342
   299     154
   300     595
   301     709
   302     156
   Length: 303, dtype: float64
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[31]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">glm_mod</span><span class="o">.</span><span class="n">fittedvalues</span> <span class="o">*</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">model</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">orig_endog</span><span class="o">.</span><span class="n">sum</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt output_prompt">
       Out[31]:</div>
   
   
   <div class="output_text output_subarea output_pyout">
   <pre>
   0      470.732584
   1      138.266178
   2      285.832629
   3      392.702917
   4       20.963146
   5     1543.545102
   6      454.209651
   7      598.497867
   8      261.720305
   9      540.687237
   10     722.479333
   11     203.583934
   12     258.167040
   13     303.902616
   14     168.330747
   ...
   288     33.470295
   289     68.855461
   290    174.264199
   291    827.377548
   292    506.242734
   293    958.896993
   294    187.988967
   295    259.823500
   296    379.553974
   297     17.656181
   298    111.464708
   299     61.037884
   300    235.517446
   301    290.952508
   302     53.312851
   Length: 303, dtype: float64
   </pre>
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>First differences: We hold all explanatory variables constant at their means and manipulate the percentage of low income households to assess its impact
   on the response variables:</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[32]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">exog</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">model</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">orig_exog</span> <span class="c"># get the dataframe</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[33]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">means25</span> <span class="o">=</span> <span class="n">exog</span><span class="o">.</span><span class="n">mean</span><span class="p">()</span>
   <span class="k">print</span><span class="p">(</span><span class="n">means25</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   Intercept                       1.000000
   LOWINC                         41.409877
   PERASIAN                        5.896335
   PERBLACK                        5.636808
   PERHISP                        34.398080
   PCTCHRT                         1.175909
   PCTYRRND                       11.611905
   PERMINTE                       14.694747
   AVYRSEXP                       14.253875
   PERMINTE:AVYRSEXP             209.018700
   AVSALK                         58.640258
   PERMINTE:AVSALK               879.979883
   AVYRSEXP:AVSALK               839.718173
   PERMINTE:AVYRSEXP:AVSALK    12585.266464
   PERSPENK                        4.320310
   PTRATIO                        22.464250
   PERSPENK:PTRATIO               96.295756
   PCTAF                          33.630593
   PERSPENK:PCTAF                147.235740
   PTRATIO:PCTAF                 747.445536
   PERSPENK:PTRATIO:PCTAF       3243.607568
   dtype: float64
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[34]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">means25</span><span class="p">[</span><span class="s">&#39;LOWINC&#39;</span><span class="p">]</span> <span class="o">=</span> <span class="n">exog</span><span class="p">[</span><span class="s">&#39;LOWINC&#39;</span><span class="p">]</span><span class="o">.</span><span class="n">quantile</span><span class="p">(</span><span class="o">.</span><span class="mi">25</span><span class="p">)</span>
   <span class="k">print</span><span class="p">(</span><span class="n">means25</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   Intercept                       1.000000
   LOWINC                         26.683040
   PERASIAN                        5.896335
   PERBLACK                        5.636808
   PERHISP                        34.398080
   PCTCHRT                         1.175909
   PCTYRRND                       11.611905
   PERMINTE                       14.694747
   AVYRSEXP                       14.253875
   PERMINTE:AVYRSEXP             209.018700
   AVSALK                         58.640258
   PERMINTE:AVSALK               879.979883
   AVYRSEXP:AVSALK               839.718173
   PERMINTE:AVYRSEXP:AVSALK    12585.266464
   PERSPENK                        4.320310
   PTRATIO                        22.464250
   PERSPENK:PTRATIO               96.295756
   PCTAF                          33.630593
   PERSPENK:PCTAF                147.235740
   PTRATIO:PCTAF                 747.445536
   PERSPENK:PTRATIO:PCTAF       3243.607568
   dtype: float64
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[35]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">means75</span> <span class="o">=</span> <span class="n">exog</span><span class="o">.</span><span class="n">mean</span><span class="p">()</span>
   <span class="n">means75</span><span class="p">[</span><span class="s">&#39;LOWINC&#39;</span><span class="p">]</span> <span class="o">=</span> <span class="n">exog</span><span class="p">[</span><span class="s">&#39;LOWINC&#39;</span><span class="p">]</span><span class="o">.</span><span class="n">quantile</span><span class="p">(</span><span class="o">.</span><span class="mi">75</span><span class="p">)</span>
   <span class="k">print</span><span class="p">(</span><span class="n">means75</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   Intercept                       1.000000
   LOWINC                         55.460075
   PERASIAN                        5.896335
   PERBLACK                        5.636808
   PERHISP                        34.398080
   PCTCHRT                         1.175909
   PCTYRRND                       11.611905
   PERMINTE                       14.694747
   AVYRSEXP                       14.253875
   PERMINTE:AVYRSEXP             209.018700
   AVSALK                         58.640258
   PERMINTE:AVSALK               879.979883
   AVYRSEXP:AVSALK               839.718173
   PERMINTE:AVYRSEXP:AVSALK    12585.266464
   PERSPENK                        4.320310
   PTRATIO                        22.464250
   PERSPENK:PTRATIO               96.295756
   PCTAF                          33.630593
   PERSPENK:PCTAF                147.235740
   PTRATIO:PCTAF                 747.445536
   PERSPENK:PTRATIO:PCTAF       3243.607568
   dtype: float64
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[36]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">resp25</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span><span class="n">means25</span><span class="p">)</span>
   <span class="n">resp75</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span><span class="n">means75</span><span class="p">)</span>
   <span class="n">diff</span> <span class="o">=</span> <span class="n">resp75</span> <span class="o">-</span> <span class="n">resp25</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>The interquartile first difference for the percentage of low income households in a school district is:</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[37]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="k">print</span><span class="p">(</span><span class="s">&quot;</span><span class="si">%2.4f%%</span><span class="s">&quot;</span> <span class="o">%</span> <span class="p">(</span><span class="n">diff</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">*</span><span class="mi">100</span><span class="p">))</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   <div class="output_subarea output_stream output_stdout output_text">
   <pre>
   -11.8863%
   
   </pre>
   </div>
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[38]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">nobs</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">nobs</span>
   <span class="n">y</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">model</span><span class="o">.</span><span class="n">endog</span>
   <span class="n">yhat</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">mu</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[39]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="kn">from</span> <span class="nn">statsmodels.graphics.api</span> <span class="kn">import</span> <span class="n">abline_plot</span>
   <span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">,</span> <span class="n">ylabel</span><span class="o">=</span><span class="s">&#39;Observed Values&#39;</span><span class="p">,</span> <span class="n">xlabel</span><span class="o">=</span><span class="s">&#39;Fitted Values&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">scatter</span><span class="p">(</span><span class="n">yhat</span><span class="p">,</span> <span class="n">y</span><span class="p">)</span>
   <span class="n">y_vs_yhat</span> <span class="o">=</span> <span class="n">sm</span><span class="o">.</span><span class="n">OLS</span><span class="p">(</span><span class="n">y</span><span class="p">,</span> <span class="n">sm</span><span class="o">.</span><span class="n">add_constant</span><span class="p">(</span><span class="n">yhat</span><span class="p">,</span> <span class="n">prepend</span><span class="o">=</span><span class="bp">True</span><span class="p">))</span><span class="o">.</span><span class="n">fit</span><span class="p">()</span>
   <span class="n">fig</span> <span class="o">=</span> <span class="n">abline_plot</span><span class="p">(</span><span class="n">model_results</span><span class="o">=</span><span class="n">y_vs_yhat</span><span class="p">,</span> <span class="n">ax</span><span class="o">=</span><span class="n">ax</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAtQAAAGyCAYAAADEaFMqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl8FPX9x/HXBImkiuUIP63iVfAuamKtJ02qJhEPFGJt
   tdrUaqO2VYFFkWJbqqFWId6tilpJa70qRWmtWeIBbapVVEBQsGrVelLjiTaIks/vj0kgWbLZa2Zn
   dvf9fDz2YbKZnfnM7vLwk8l7P1/HzBARERERkfQUBV2AiIiIiEguU0MtIiIiIpIBNdQiIiIiIhlQ
   Qy0iIiIikgE11CIiIiIiGVBDLSIiIiKSAV8basdxfus4zmrHcZb3sc01juO84DjOMsdxyvysR0RE
   RETEa35fob4VODLeDx3HOQoYaWa7APXA9T7XIyIiIiLiKV8bajP7O/B+H5uMBZo6t30cGOQ4ztZ+
   1iQiIiIi4qWgM9TbAa91+/51YHhAtYiIiIiIpCzohhrAiflea6GLiIiISM7YLODjvwFs3+374Z33
   9eA4jppsEREREckKM4u94NunoBvq+cCPgTsdxzkQ+MDMVve2oZl6aulp+vTpTJ8+PegyJGT0vpDe
   6H0hvdH7QnrjOCn10oDPDbXjOHcAFUCp4zivAT8H+gOY2Y1m9lfHcY5yHOdF4BPgND/rERERERHx
   mq8NtZmdlMQ2P/azBhERERERP4XhQ4kiaamsrAy6BAkhvS+kN3pfSG/0vhCvOLmQTXYcx3KhThER
   ERHJTQ8/DBdfDIsWOSl/KFFXqEVERESkYL3wAhx/PJxxBvw4zSCyGmoRERERKTgffACTJ8NBB7m3
   556DE05Ib19qqEVERESkYHz+OdxwA+y+u9tUr1gBU6bAgAHp7zPoOdQiIiIiIlnx4IMwcSIMHQoP
   PABlZd7sVw21iIiIiOS1F16ASASefRZmzoRx4yCN9VviUuRDRERERPLSBx+4jfRBB8Ghh7o56fHj
   vW2mQQ21iIiIiOSZzz+H66+H3XaDNWvcK9MXXACbb+7P8RT5EBEREZG8sWABTJoEw4ZBNAr77uv/
   MdVQi4iIiEjOe/55dwzeypUwaxYcd5z30Y54FPkQERERkZz1/vvu5I5DDoGKCjfecfzx2WumQQ21
   iIiIiOSgzz+HX//anSfd3u5+4HDyZP9y0n1R5ENEREREcko06uakt9kGWlpg772DrUcNtYiIiIjk
   hFWr3DF4//oXNDbCscdmN9oRjyIfIiIiIhJq770H550Ho0fD4Ye7OemxY8PRTIMaahEREREJqc8+
   g2uvdXPS69a5OelJk6C4OOjKelLkQ0RERERC54EH3OZ5+HB46CEYNSroiuJTQy0iIiIiofHcc25O
   +qWX4Ior4OijwxPtiEeRDxEREREJ3LvvwrnnurOka2pgxQo45pjwN9OghlpEREREAvTZZ3D11W5O
   ev16d6XDCRPCl5PuixpqERERkZCIRqNUV9dSXV1LNBoNuhxfmcH997vZ6Pvvh0cecRdqKS0NurLU
   OWYWdA0JOY5juVCniIiISLqi0SjjxtXR3n4ZACUlU5g3r4mampqAK/Pes8+6Hzh85RU3J33UUeGJ
   djiOg5mlVI2uUIuIiIiEQGPj7M5mug5wG+vGxtlBl+Wptjb48Y+hstJtolesyI0PHSaihlpERERE
   fLVuHVx1Feyxh9s8r1rlLtTSv3/QlXlDY/NEREREQiASqae1tY72dvf7kpIpRCJNwRaVoa6cdCQC
   O+8MixbBnnsGXZX3lKEWERERCYloNLoh5hGJ1Od0fnrFCjcn/dprbk56zJigK0pOOhlqNdQiIiIi
   4pl33oGf/xzuuQd++lM466zcinboQ4kiIiIiEoh169wr0Xvu6TbQq1bBOefkVjOdLmWoRURERCRt
   ZvDnP7s56V13hb/9zf3wYSFRQy0iIiIiaVm+HCZOhDffhGuvhSOPDLqiYCjyISIiIiIp+e9/3Wz0
   4YfDuHGwbFnhNtOghlpEREREkvTppzBrlpuTHjAAnn8efvSjwshJ90WRDxERERHpkxncdx9Mngy7
   7w6tre5/xaWGWkRERETieuYZmDABVq+GX/8acng0tm8U+RARERGRTfz3v3DmmVBVBSec4Oak1Uz3
   Tg21iIiISAhFo1Gqq2spL6+kvPxQqqtriUajvh/3009h5kw3J73FFu486R/+EDZTriEuPTUiIiIi
   IRONRhk3ro729ss675kMHEhrax3z5jX5siS5Gdx7r5uT3msvePRRd660JKalx0VERERCprq6lpaW
   sUBd5z1NwHxgLFVV81mwYK6nx1u61J0n3dbmrnZYVeXp7nOKlh4XERGRUOiKK2QrppCuRHXmynmk
   a/Vq+MEP3BnS3/oWLFlS2M10uhT5EBEREU/FxhX8jClkIlGdQZ5HJFJPa2sd7e1d90wG6igpmUIk
   0pTx/teuhauvdrPS3/uem5MeNCjj3RYsRT5ERETEU73FFfyIKWQqUZ1Bn0c0GqWxcTZtbe8Cn1Na
   ujWRSH1GDb0Z/OlPcP75sPfebkO9yy7e1ZwP0ol86Aq1iIiISAjV1NR4ejV8yRJ3nvT778NNN7nL
   hos31FCLiIiIp2LjCl7FFLyWqM5cOY9E3n4bpk2D+++Hiy+G00+Hfv2Criq/KPIhIiIinuuKKwAZ
   xxT8lKjOXDmP3qxdC1deCbNmwfe/DxddBF/8YtBVhV86kQ811CIiIiJ5xAzuuQcuuAD23dfNSY8c
   GXRVuUMZahEREZEC9tRTbk56zRr47W/hG98IuqLCoDnUIiIiIjnurbfcWMcxx0BdndtYq5nOHjXU
   IiIiIjmqvR1mzICvfAWGDYPnn4czztCHDrNNkQ8RERGRHGMGd98NU6bAfvvBE0/AiBFBV1W41FCL
   iIiI5JAnn3Rz0p98AnPmQGVl0BWJIh8iIiIiOeDNN91lwo89Fk47zW2s1UyHgxpqERERkRBrb4eG
   Bhg1Cr70JTcnrcVZwkWRDxEREZEQMoO77nJz0l/7GixeDF/+ctBVSW/UUIuIiIiEzBNPuDnptWvh
   97+Hr3896IqkL4p8iIiIiITEG2/Ad78Lxx8PP/iBe1VazXT4qaEWERERCdj//gcXXwx77w3bb+/m
   pE87TTnpXKHIh4iIiEhAzOCOO+DCC+Ggg9wVDnfaKeiqJFVqqEVERCTnRaNRGhtnAxCJ1FNTUxNw
   RYk9/ribk/7sM/jDH2D06KArknQp8iEiIiI5LRqNMm5cHS0tY2lpGcu4cXVEo9HAaqmurqW6ujZu
   Da+9BqecAuPHw1lnuR9AVDOd23SFWkRERHJaY+Ns2tsvA+oAd25zY+PsrF+l7mrs3VqgtbWOefOa
   NtTxyScwcyZcey2cfbabk95yy6yWKD5RQy0iIiLigXiNfVVVDbffDlOnwiGHwNNPw447BlureEsN
   tYiIiOS0SKSe1tY62tvd70tKphCJNAVbVKcPPtiFgw+G9evhzjvdhlryj2NmQdeQkOM4lgt1ioiI
   SDDC8KHEnpGPL9CvXxGDBx9NY+MATjkFivTJtZzgOA5m5qT0mFxoVNVQi4iISC64774HmTTpPV57
   bQzf/OZ/ufHGEcpJ55h0GmpFPkREREQy1NHhjr6bOvUIvv51eOQR2GGHgUGXJVmiPz6IiIhI0pIZ
   C5fJ9rno0UfhwAPhuuvg7rvh9tthhx2CrkqySZEPERERSUrsWLiSkik9xsJlun2uefVVd4XD1la4
   9FI4+WTlpPNBOpEPX192x3GOdBxnleM4LziOM6WXn5c6jtPsOM5Sx3FWOI7zPT/rERERkfT1HAvn
   NspdHwT0Yvtc8fHH8NOfQnk57LYbrFqFPnRY4Hx76R3H6QdcBxwJ7Amc5DjOHjGb/RhYYmb7ApVA
   o+M4ynWLiIhI6HR0QFOT20S//DIsXQrTp8MWWwRdmQTNz+b1a8CLZvYKgOM4dwLHASu7bfMWsHfn
   11sB75rZ5z7WJCIiImlKdd5zmOdDp6q1FSZOhH79YO5cNzMt0sW3DLXjOCcANWb2g87vTwEOMLNz
   um1TBDwM7AoMBE40swd62Zcy1CIiIiGQ6rznMMyHzsQrr8CUKfDYY/CrX8G3v61oR74L1Rxqx3Fq
   gSMTNNQXAaVmNsFxnBFAC7CPma2J2ZcaahEREcmaNWvcBvqGG+Dcc+H88+ELXwi6KsmGsM2hfgPY
   vtv32wOvx2xzMDADwMxechznZWA34MnYnU2fPn3D15WVlVRWVnpbrYiIiBS8rpz0tGlw+OGwbBkM
   Hx50VeKnhQsXsnDhwoz24ecV6s2A54HDgTeBJ4CTzGxlt22uAD40s184jrM18BSwt5m9F7MvXaEW
   ERHpQ65HK8Lg73+HCROguBiuugoOOCDoiiQIoYp8ADiOMwa4CugH3GJmlzqOcyaAmd3oOE4pcCuw
   A+7EkUvN7PZe9qOGWkREJI58n/fst5dfdnPSjz8Ol10G3/oWOCm1U5JPQtdQe0UNtYiISHzV1bW0
   tIzFnfcM0ERV1XwWLJgbZFmht2YN/PKXMHu2e2U6ElFOWkK4sIuIiIhId2FYinz9evjtb9150m++
   Cc884y7UomZa0qVFVERERHJcNuY9e5HRjo2mtLbWZT2asmiRO0+6pATuuw/23z9rh5Y8psiHiIhI
   HvDzQ4leZbSDjKb8+99wwQWweDFcfjmceKJy0tK7sI3NExERkSypqanx7UpvY+PszmbabYTb2937
   cuFDjx995Oakb7oJJk2C3//evTot4iU11CIiItJD7NVur2RzKfL16+HWW91s9JFHwvLlsO22vhxK
   RA21iIiIV/JhFnRvOedp086htXVKxo1wTU0N8+Y1dXuO/MlPL1zoTu0YOBD+/Gf46lc9P4RID8pQ
   i4iIeCBfZkHHyzlHIvWh/2XhpZfcJcKXLHFz0iecoJy0pE4ZahERkYDkcs44GX5mtDP14YcwYwbc
   cgtMngy33w4DBgRdlRQSzaEWEREJoaDmNUci9ZSUTAGagKbOeId3OWovrV/vLsqy227w7ruwYgVM
   napmWrJPkQ8REREPeBn5SGZfveW1vcpw50IW/OGH3XnSX/wiXHUVlJcHXZHkCy09LiIiEiCvGtFE
   85p7a7inTTuHGTOuzfkMdyIvvODmpJ95BmbOhPHjlZMWbylDLSIiEqBs5Yx7y2tfccUleZ3h/uAD
   aGiAOXPcnPSddyraIeGhDLWIiGRNULngXJNLOeZY6bzGfT3m88/hxhth993dpnrFCrjwQjXTEjJm
   FvqbW6aIiOSy5uZmKynZ2mCOwRwrKdnampubgy4rEM3NzVZVNd6qqsbHfQ762qa357KhoSHw5zed
   17ivxzz4oNmoUWYVFWZPP52FExAxs86+M6VeVRlqERHJikS54ELh1YcX/fxQYrrSeY17e8zBBy+m
   tPQ6Vqxwc9LjxiknLdmjDLWIiEjIeTWvure8dphnRSfvqyxePI6GBrj7bth886DrEUlMGWoREcmK
   sOaCcy3XHeZ603mNI5F6+ve/EPgn0E5R0VP87ndPccEFaqYld+gKtYiIZEVNTQ3z5jV1iyQEP9It
   Nn7R2lrn+6i5SKSe1tY62tvd792msympxwZRbyrSeY2ffnoon3/+MPA/4Ew22+zPDB58p//FinhI
   GWoRESlYQeW6080651MO/fnn3fF3Dz30Fu3tK4HDOn+Su+ck+UEZahERkRyQH1nn9Lz/Plx8Mfz+
   9+74u/b2CTz00FFBlyWSETXUIiJSsDKJX6TCq+kb2arXD13zpC++2J3a8dxz8H//B6NGfZ9HH83N
   cxLposiHiIgUNL9HzXk1Jq/7/oIcjZeOaBQmTYJttoErr4S99479ee6dk+SvdCIfaqhFRER8lE+5
   51StWgWRCPzrX9DYCMceq3nSEn7pNNQamyciIiKeeu89OO88OPRQOOwwePZZGDtWzbTkLzXUIiIi
   Pgrr/G0/fPYZXHst7L47rFsHK1e6V6iLi4OuTMRfinyIiIj4rBAyws3Nbk56u+3giitg1KigKxJJ
   jzLUIiIiklVdV6FffNFtpI8+WtEOyW3KUIuIiEhWvPsunHsufP3rUFUFK1bAMceomZbCpIZaRERE
   kvbZZ3DNNbDHHrB+vXuFeuLExDnpaDRKdXUt1dW1RKPR7BQrkiVa2EVEREQSMoMHHnBz0jvsAA8/
   DF/5SnKPjZ3F3dpal9EsbpGwUYZaRERE+vTss25O+uWX3Zz0UUelFu0o5FncknuUoRYRERHPtLXB
   j38MlZUwZoybk9aHDkU2pYZaRETynvK7qVm3Dq66ys1JO4674uF550H//untr5BmcUthUuRDRETy
   Wmx+t6RkivK7cZjB/fe78Y6dd3bjHXvu6c2+C2EWt+QHzaEWERGJofxuclascD9w+NprbiM9ZkzQ
   FYkEQxlqERERSck778APfwiHHQbHHgvPPKNmWiRVaqhFRCSv5Xt+N918+Lp1GyMd/fu7Oelzzkk/
   Jy1SyBT5EBGRvJev+d108uFm8Je/uDnpXXaBWbPcDx+KiEsZahERkQKSaj58+XJ3VcM333SvTh95
   ZNZKFckZylCLiEioaXydH5YDtZ235Rvu7f5c33XXI5x1Fhx+OBx/PCxbpmZaxEtaelxERLJCy09v
   5FUEpaKinJaWy4FrOu85l4qKC7o91zOB3Wlp+TLHH/8qq1btyJAhnpyCiHSjhlpERLKisXF2ZzPt
   xhPa2937Cq2h9vIXi0WLnsZtpuu63TefhQufpr39duCwznvn8ckntzFkiEYFivhBDbWIiEgW+f2L
   xZo1O/Kvf30P2KHbvR95sm8R6Z0y1CIikhWpjK8LQ9Y6toZ0a/LzXHo+p3fRr5/DqlWXctJJmzFg
   wF50f64rKsoDf05F8paZhf7mlikiIrmuubnZqqrGW1XVeGtubo67TUnJ1gZzDOZYScnWcbf1s87u
   NRQXD7Pi4kEp19TbuTQ0NHh6fvPnR22XXZqsf/8Pbdy4l+299zYeu+u59vqYIvmss+9MqVfV2DwR
   EQmVMCwV3lsNcAPwWEo1xTuXSKQ+4w8lmsG998LkybDXXu486V13Ta0OLb8usql0xuYpQy0iInHl
   64IoQaupqcnouVy61J0n/c47cMMNUFXlYXEikjI11CIi0qtsjrnr3rhXVJTT2jqF9nb3Z27Wuint
   fc+YMYMrrrgVgGOPPZQ331wD9P0LQiRST2tr3YYaiovPBz5j3bqmlGqK3U+m57J6NVx0EcyfD7/4
   BZxxBmyWxP/Jva5DRHpS5ENERHqVrZhAb8tnT5t2TudIuMyujM+YMYOLLuo5pxl+AIxKuEx37NV5
   IK2r9V5c5V+7Fq6+GmbOhO99z22qBw1KbR/6a4NIcrT0uIiIeCZbDbWfxxk6dCTvvfdTemah5wNz
   PT2OX8zgT3+C88+Hvfd2G+pddgm6KpH8pgy1iIh4RjGBYC1ZAhMmwHvvwU03ucuGi0g4aQ61iIj0
   qqamhnnz3Ku4VVXzfctP95ylPJmioghtbe96Mit50qTTcGMeTZ23c4GdSTQHO0hvvw2nnw5jxsDJ
   J7uNtZppkXBT5ENERAIXjUaZOvUSli17jo6OKwESZpyTlc6HEoOwdi1ceaU7/u7733dz0l/8YtBV
   iRQeZahFRCRnFeqsZDOYO9fNSe+7r5uTHjky6KpECpcy1CIiIjnkqafcedIffgi33AKHHRZ0RSKS
   DmWoRUQ8FI1Gqa6upbq61pMMcCHpmaUOb8bZC2+95cY6jjkGTj0Vnn46s2Za7zuRYCnyISLikd7m
   Kfv1Qb58le+zktvb3Zx0Y6O7KMtPfpJ5TlrvOxFvKUMtIhKgQs0AS2Jm8Mc/wgUXwH77weWXw4gR
   3uxb7zsRbylDLSIiEjJPPunOk/74Y5gzByorg65IRLymhlpExCNaCCW3+BUv6drv2rWD2XzzX7Bi
   xXY0NLhLhvfr58khetD7TiR4inyIiHgo3zPA+cKv3HE0GuX4489k7dp5wB5sttls7rprT8aPP8KD
   qvs+rt53It5QhlpERCQJfuSOzWCffa5g+fJ6YEvP9isi2ZVOQ62xeSIikhGNbIMnnoBDD4VXXjke
   aA26HBHJMjXUIiKStq7oREvLWFpaxjJuXF1ONNVezbx+4w347nfh+OPh9NPhzjtfoKTke33uV7+A
   iOQfRT5ERCRtuTyyLZPc8f/+B7NmwdVXw5lnwtSpMHBg4v1qZrRI+GlsnoiISJJqampSbmTN4M47
   YcoUOOggd+nwnXZKfr+NjbM7m2n3F5D2dvc+NdQiuU0NtYiIpK2QRrY9/rg7T3rdOvjDH2D06KAr
   EpGwUIZaRETSVlNTw7x5bsyjqmp+zsUXkskzv/YanHIKjB/vxjsWL06/mfYquy0i4aIMtYiIFKRE
   eeb//Q9mzoRrroGzz4YLL4Qtt+xrj8kfVzOjRcIrdHOoHcc5ErgK6AfcbGaX9bJNJXAl0B9oM7PK
   XrZRQy0iIp6K94HK5ua53HGH20AfcghcdhnsuGOQlYpINoVqDrXjOP2A64AjgT2BkxzH2SNmm0HA
   r4FjzewrwAl+1SMiIuEWhnFyH3ywCwcfDFdd5X748M47c6OZDsNzJ97Qa5mbEl6hdhxnS6DdzNY7
   jrMbsBvwgJl9luBxBwE/N7MjO7+/EMDMftVtmx8C25jZzxLsS1eoRUTyWBDj5Hoe8wv061fE4MFH
   09g4gFNOgaIc+ZSRRvHlD72W4eDXFeq/AZs7jrMdEAVOBeYk8bjtgNe6ff96533d7QIMcRznEcdx
   nnQc59Qk9isiInmm5zg5t6Hoyhn7paamhjvuuI0vf7mE/v2P5Fvf2peXXx7Ad7+bO800BPPciT/0
   WuauZMbmOWb2P8dxTgd+Y2aXO46zLInHJXNJuT9QDhwOfAF4zHGcf5rZC7EbTp8+fcPXlZWVVFZW
   JrF7ERGRTXV0uKPvpk49gq9/HR55BHbYYWDQZYlIABYuXMjChQsz2kdSc6g74xvfAU7vvCuZ393f
   ALbv9v32uFepu3sN94OI7UC74zh/A/YB+myoRUQkvlycIpHNedaPPurOkwa4+244+ODU9xGm5zjR
   cxemWqVvhTTXPUxiL9T+4he/SH0nZtbnDagA5gNTOr8fAVyTxOM2A14CdgKKgaXAHjHb7A48iDsF
   5AvAcmDPXvZlIiKSWHNzs5WUbG0wx2COlZRsbc3NzUGXlZTm5marqhpvVVXjfan51VfNvv1ts+22
   M/v9783Wr09vP2F8juM9d2GsVfrm978DSayz70zYI3e/JT02z3GcLczsk1SadcdxxrBxbN4tZnap
   4zhndnbIN3ZuMxk4DegAbjKza3rZjyVbp4hIIYs3Cm7BgrlBlhWojz92R9/95jdwzjlw/vmwxRbp
   76+8/FCWLFkPbAvUA2+H9jnW+0Ekdel8KDFh5MNxnIOBm4GBwPaO4+wL1JvZDxM91sweAB6Iue/G
   mO9nAbNSKVpERCSRjg74/e9h2jSorISlS2H77RM+rE/RaJRly57DXT4B3Eb1lMx2KiI5L5ks9FW4
   s6TbAMxsKW4MREQkb4V5FmxftfW1tHU0GqW8vJKhQ0dSXn5o6M7LS62tcMABcP31cM89cNttmTfT
   4E5h6Oi4kq4pDHAZRUVz+lw+PMj3kpY6F8mSRJkQ4InO/y7pdt+yVLMlmdxQhlpEsihsudPumcqG
   hoaEtfWWwWxubrbi4mEbHgelVlw8KO8ymq+8YnbiiWbDh5vddlv6Oel4qqrGdz5/1nmbY2VlFXG3
   D8N7SZlckdSQRoY6mWb2HuAQYAnuhwsnA3emeqBMbmqoRSSbemuaqqrGB1JLbENWVDQ0rdp6Oyc4
   MLDz8tqaNWbTppkNGWI2fbrZxx/7c5xUG+QwvZdEJDnpNNTJjM07G7gad1GWN4AFwI88u0QuIiJx
   9VzoATo6bgi2oJDp6IDf/c7NSR92GCxbBsOH+3e8mpoa5s1r6jaGTqvYiUgSH0o0s3eAk7NQi4hI
   KHg9C9bbOcCHUFQ0kY6O1GqLROpZtOhU1q3rumcyjrOWtrZ+RKPRtGry6rzS3U9j4+P89KeDWLfu
   f4wYcT2nnFLL8OH+N7c1NTVJ16i5wiIFItElbODWmNtvgd+meik8kxuKfIhIlnmVO800Q9vb4xsa
   GtKqrbm52crKKmzgwB3McQYaRNLO9TY0NFhR0eCMs8HpPD///rfZ6NFvGbxmsCj0mXBlmEVyC37M
   oXYc5wQ2LiNeAowD3jSzc3zp8HuvwRLVKSISRl7MAfZ6pbtMa4pGoxx11Hfo6GhMex/p1LJmDVx6
   Kdx4IwwefAcvvbSejSPrmoAbqKraVjOWRSQjvsyhNrN7Yg5yO/CPFGsTEZE0pRIxyAZ3dNwuWTve
   +vXQ1AQXXQRVVfDMM3Daaffw0ktjs1aDiEhfkvlQYqxdgWFeFyIiko/CmKH1pqZDgCkbvisqmkgk
   cgeQ2hX1RLUsWgQTJ0JJCdx3H+y/v7v/trbVOM4ENv7xcjLFxZ8TiUxP8TxIuWYRkU0kyoQAHwNr
   Om8fAS8AtalmSzK5oQy1iOSwMGZoM6lpY+45YnCgFRUNtYaGhpifJZ+J7q2Wl14yq60122EHszvv
   NOvoiD32HIOIOc5gGzhwBysrOySjbHnQs6JFJDzwI0MdBspQi4iES7wrupnmsz/6CH75S7jpJpg0
   yb2VlGz8ebz9RyL1jBtX1zli0L3SPW9eciPtvMi5i0j+SCdDHXfpccdx9nMcpzzeLfNyRaSQhHkp
   73w1Y8YMhg4dydChI5kxY0av26T7utTU1LBgwVwWLJibUTyiq8YhQ3Zl3Lj72W03WL0ali93Z0t3
   b6b70nNet9tYdzX8IiK+i3fpGlgIPBLvluql8ExuKPIhktP0J/Xsa2hoMNiq21LjW22IZXTp63VJ
   NxKSymtdV1fXWeMDBu8aPGpnn31Lr/tMtPR6JisSxsZIioqGWllZhd6jIgUKP5YeD8NNDbVIbtPy
   y9k3ZMiITZ7zIUNG9Ngm3uvixezsRM14c3OzQZnBK33WmOwcbi9qLis7xJPZ2iKS29JpqJOa8uE4
   zihgD2BRradAAAAgAElEQVRAtyvbv/PqKrmIiIRH7HLn7e3ufclGOxKN+fvoIzjrrP8BDwP/BnZM
   qZZFizbNN2e6JHhNTU3nOMAfkO55i0jhSthQO44zHagA9gLuB8YArYAaahFJShhHx+W7SZNO46KL
   zu12z7lMmnRBj21iX5eioom0te1JehNVE1u/Hn77W/jZz6CoaCvgauB64LIeNXb/wGNb2+pN9tPW
   tprq6toN59DV8IZtXreIFJBEl7CBFUA/YFnn91sDD6Z6KTyTG4p8iOS8MI6Oy3cNDQ02ZMgIGzJk
   xCb56S5dy5EXFQ3dsBR5cfEgKy4ellH0Ifb1fughs733Nhs92uzJJzcdvQeDra6ubpPoRnHxMCsu
   HtTt+8xr66vmMGT99W9FJFj4tPT4YjPb33Gcp4DDcGdRrzKz3fxr8zepwRLVKSIi6eltbFxZ2U2U
   lm4NpL7QSTQa7TbC7k1gb4qLyzn//NVccsm+OM7G7WJH7/Vey62Ulg4FoK3tXZYsOQ2/RtwFvcBL
   z+cutfF/IuINX5YeB550HGcwcBPwJPAJ8Gga9YmISI4oLd067SbVzT1fARwEbAW8wrp1D3LFFecz
   enRTyhGN0tKhG2rpinr4JejYSKb5dREJRl9zqH/jOM6hZna2mb1vZjcA1UCdmZ2WvRJFRLIvl+dm
   d9VeXl5JefmhCc8hEqmnpGQK0AQ0dWbc61M6VtcxPv8cXn+9ChgHvAQ8COwHnNpjNnS85zdRLZnU
   KiLim3hZEGAC8BjwKnA5UJZqnsSrG8pQi0gWhSVLm47Y2qHUIJLwHNLJ7W6ad661nXb6yEaNetc2
   3/zwzmx06mP5EtWSzxnjXH7vieQLfMpQ7wR8G/gW8AXgduAOM/uXTz1+bzVYojpFRLySy0tR91Y7
   zAfGen4Omx5rDXvvfT1Ll17AggVRpk69hGXLnqOj40pgYx64sXF2zj6/2RB0jluk0Hm69HgXM3vF
   zH5lZmW4jfU4YGWaNYqISJ747LMvAF/tds+9bL314ziOm0V++ulW/vrXO6iqmk9V1Xx9uC5JXi3r
   LiLZk7ChdhxnM8dxxjqOczvQDKwCxvtemYhIQMKS042XM45Go5SXVzJw4LZstdUOlJdXbvh5bO0w
   GdjZ03P4/HO4/npYuvRm+vV7GrgL93k6f5Nj9NYcevX85nLOXUTyS9zIh+M41bhXpI8GngDuAOab
   2cfZK29DLYp8iEhWBfln92g0flwCYOzYU1m37ru4DeksAIqLz2f+/N8DMHXqJbz66tsMHjyQrbba
   gtLSrT07h5YWmDgRhg2DK6+E1avTe54yfX41Xk5E/JJO5KOvhvph3CZ6rpm950F9aVNDLSKFYmOj
   uDNwFrE5Y6Azf+zmomPnNa9atcqzJrN703viiRO4777RrFwJs2bBccexYZ50EHI55y4i4ebpHGoz
   OyzzkkREJBUb5xDPT/mxr776umczjDc29lcCe9PSsg1nnPE899yzG5tvnvLuRETyWjILu4iIFJzY
   SAKQ5QhIPRuvvtKZM3YjH4sWdUU+Jm/4eXHx+ey4466859HfE2fOvJn29nuBAzvvuYtXX72bzTfv
   /QpwtiMykUg9ra11tLe733d/fkREsi7VOXtB3NAcahHJok3nKw+z4uJBWZkN3PPYESsqGmplZRWb
   zGkuK6uwLbf8kg0cuP2Gn3s1w7i52WyLLf5j8OYmM6QT15y92cn5PI9aRIKDH3Oow0AZahHJpt5n
   Od+Au9aV+72fed1MrvZm8thVqyASgeefh1NOeZrLLjuKtWsT57GVZxaRfOJphtpxnI+BeF2smdlW
   qRxIRESSU1NTk3ZkIp3HvvceXHwx/OEPcOGF8Kc/weabl3PQQU3dmnNN0BARiaevDyVuCeA4TgPw
   JnBb54++A2zrf2kiIsGIzecWF58PfMa6dW5GN15eN4wr3PVV02efwY03us10bS0895w7Dq9Lss25
   8swiUuiSWXr8GTPbO9F9flLkQ0SyLdUPJYZxLnJfNTU3w6RJsO227jzpUaMyP1amv0yE8RcSESk8
   ns6h7rbTx4Bf486kBnexlx+Z2cFpVZkGNdQiEnZhzBH3VtNBBz3FoEHX8OKL0NgIxxwT7DzpLmH8
   hUREClM6DXXCpceBk4ETgdWdtxM77xMRkSSEZ4nsr/Hkk5dQVQUrVsCxx4ajmYbu87frALex7rpa
   LSISdgkbajN72czGmllp5+04M3slC7WJiKQliAY2EqmnpGQK7kSQps4ccf2GK68tLTvT0vImY8Z8
   k5Ejy3rUlk69fT2m62dtbavp338q8DjQTr9+/+S2255i4kQoLvb09JOuLd62Tz21zL+CRET8lmiu
   HrAb8BDwbOf3ewMXpTqfL5MbmkMtIkkKaiZy17Fj5yJXVY03iBhs3fnf0h61NTQ0pFxvX+fY82ct
   Bs/bwIFP2EEHnZe12dDJnk/szG3YKpDXTUSkO9KYQ51MM/s34ABgSef3Tldzna2bGmoRSZbbwM5J
   akGS7NVzYGdNm9Y2ZMiIlOvt6xzdn83r9rMWO+KI7J1/Ks//pttGbMiQEb4v1KIFYUSkL+k01Mlk
   qL9gZo93u6JtwGfeXB8XEclvkUg9RUUvZOVYbW2watUZwJHd7n0jNDnpxEax3377bPggpx+xnY0R
   nLG0tIxl3Li6gHPtIpIXEnXcwAPASDZeoT4BeCDVzj2TG7pCLSJJCjLyEU9DQ4MVFQ32LfIxYMBw
   O/PMlVZaajZ27Cs2YMDugZ1/+pGPjdv6+RoG/RcMXR0XCT98inyMwM1Qt+Mu8PIPYKdUD5TJTQ21
   iKQijE1LV01lZYdYWVlFj9rSqbe5udmOOGK87bvvDNtuu4+tpsbs2Wd7Hiuo80/l+PFz5/40vUE2
   1GH8ZU9ENpVOQ53MHOp+ZrbecZwtgSIz+8j76+R90xxqEZGenn0WJk6E//wHrrgCxowJzwi8TPk5
   0zvIeddhnFUuIpvyaw71y47jzMb9YOKatCoTEekmPHOZg5fqc9HWBsce+x/Kyj7kP/+5hcbGBRx1
   VP400xB/BKEXampqmDfPbWSrquZr8RgR8UQyV6i3AI7BXSGxHPgzcJeZ/d3/8jbUoCvUInkil1bE
   83sp7FSei3Xr4Ne/hl/8Yh0ff9zE+vVbAutC/fxlwo/nPuilzXPpvS9SyHxZejzmAIOBa4CTzaxf
   ivWlTQ21SP7IlT97Z6P5Sea5MIO//AUiERg5Ej744Fwee2y/Ph8jmwpLMxt0Uy8iiaXTUG+WxE4d
   oAL4Fu4spsW4y4+LiATG78ak51LY0N7u3pfNBmj5cpg0Cd54A66+2s1JV1e/AeyXtRryRRheT3Aj
   J2qiRfJPUhlqYALuAi+jzOxEM9OlEJEC5EX22Yt8bLKzhDNZ0ttdCnt5SjVVV9dSXn4o5eWVSR0z
   3nPxzjtw9tlw+OFw3HGwbJnbTPf1GBERCVBfI0CAfsDPUh0d4vUNjc0TCZyXI78yHevW2+iz2BX2
   0qk39jHuUtiRFOYpbzpnOpWxcX/+c9RmzTIrLTU77zyzd99N/JjY/Qc9Mi8ZQdSokXUikix8mkO9
   ONWden1TQy0SvKAXxEhUS9fy3l2NUjr1JtOo9/249J6jjg6ze+81GznS7OijzVauTPkpMbPcaBqD
   rDEXftkQkeCl01AnzFADrY7jXAfcBXzS7cr2095eKxcRSU4kUk9rax3t7V33TAZuA2o2ZGO90n0p
   bD8884w7T/rtt+G66yA2XptKVjwsOeG+BFmj8ssi4pdkMtRlwF7AxUBjt5uIFJAwZXe7zxIeMuQS
   3OasZ6OUar3RaJS2ttUUFU3s8ZiKivKEOeyNx9oZt7lPfMz//hfOPBOqqqC21s1J99ZMJ5MV90NY
   Z4WHtS4RKXCpXtIO4oYiHyKhEMY/mfcVIUi23p77iFhR0VArK6uwhoaGpOMJfS0t3t3atWaXX242
   dKjZxIlm770X/9xSja14FafwM5bR0NDQmU3fmFNvaGgIvC4RkS74tPT4NsAMYDszO9JxnD2Bg8zs
   Ft+7/Y01WKI6RaRwZTpCL948aMCzmdlmcO+9MHky7LknNDbCrrumV1dfx/dinKCfs8Ldfe+MO0AK
   YGeqql5Oat+5MsNcRHKbL3OogTnArcC0zu9fAO4GstZQi4j0JezZ2KVL3Zz0O+/ADTe4MY9kxGbF
   3QhJU5+PCftz4RoFzOr8uomNzbWISG5KJkNdamZ3AesBzOwz4HNfqxIRXymH2lO8vHXs/cXF59PW
   tjrp5+3OOx9h+PAWDjjgfUaNepalS/tupmNfl+5Z8bKym9h22y9x8sk/orz80JRet1Rfbz/z8pns
   O51cvN7nIpIViTIhwEJgKLCk8/sDgUWpZksyuaEMtYhnlEPtXby89cZsdIUVFw9K6nlrbzf7/vef
   N2gzWGHwhw3b93WcvrLgxcXDuuWOS624eFBSr1u6r7efeflM9p1eLl7vcxFJHj7Nod4PeBT4sPO/
   LwD7pHqgTG5qqEW8E6Z50rkkmeeto8PsnnvMdt7ZbNiwxw3m9ti+rOyQuE1eX/uPN3c7mdetUF/v
   Qj1vEclcOg11wsiHmT0FVACHAPXAnma2zMOL5CIivsrGn/6XLIFvfAOmT4fZs2HffS8D1vTY5tVX
   3+42g7mO9vbLPJ2ZLSIiwUjYUDuOcyJQYmYrgHHAXY7jlPtemYj4IkzzpLPBq1nO8Z63t9+GM86A
   MWNg772fZeutv8nll9dSUVG+yfY77jg8pf13zcBua1vNZptFNvwMJlNcvIpIpD7hLwupvN75lDku
   tPe5iAQs0SVsYHnnfw/FzVMfAzyR6qXwTG4o8iHiqTDOk/aLl3/67/68zZ+/wC691J0nPXmy2dy5
   D24S52hoaOjxPCfK9Xbff+wM7OLiQTZixL42ZMgIKys7JKn99bbfvuZo51vmuJDe5yLiHXzKUC/t
   /O+vgO90fr0k1QNlclNDLSLp8jpL29Fh9sc/mu20k9nxx5u98EJqx0m2yUtmf16emzLHIiKudBrq
   ZOZQv+E4zmygCviV4zgDSG7cnohI4NKZ5RzP00/DhAnw4Ydwyy1w2GGp7yM35kSLiEgqkmmMTwSi
   QLWZfQAMBs73tSoRySt+ZXOT2W/3Wc5VVfOZN6+pz4a2t32+9RZ8//tw9NFw6qluYx3bTKeT2e2r
   /mT252VOWJljEZEMJHMZG3d03nnAOUB5qpfBM72hyIdIzvIrm+vHfmP3OWDAjva97z1vQ4eaXXCB
   2QcfJH58spndZOpPNvvsVU5YmWMRkfQiH477uPgcx/kZ8E3gT4ADHAfcY2aX+Nfmb1KDJapTRMKp
   urqWlpaxuKPiANyrxQsWzA3dfjfd58f83/+t4NFHD2TEiIzKTeJY3jwvIiKSGcdxMDMnlcckE/k4
   BdjfzH5uZj/DXSnx1HQKFBHxXhSoBW6grW21B/sb2u3rVvbZZ6bnzXQuy6fReiIiXkmmoX4DKOn2
   /QDgdX/KEZF80dV4tbWtprj4fLzO5kYi9RQXT8D9nX8scBbPPvuvtJu8N98Es2uA/YF/AL+jpOR7
   vuWIs51Z9qIR9mqmt4hIvokb+XAc59rOL7cHvgYs6Py+CncO9Tj/y9tQiyIfIjmkq/FyVwWE4uIJ
   7LXXPpSWDiUSqfdsykV5eSVLlpxGJrGJ9nZobIQrr4T6eth//we54YbrATyttTfRaHTDSol+Hiv2
   9SgpmdLrhzMT1aOYiogUgnQiH32NzXsKsM7/3tv5tYO7uEtS3a3jOEcCVwH9gJvN7LI42+0PPAac
   aGZ/SrZ4kUKSrebLC42Ns7stsQ3r1kFpqfeNV2np0MQbxWEGd90FU6bA/vvD4sXw5S8DHMH48Ud4
   VmNfsjVCL/b1cH+JmN3j2LFNd2trXcKJKCIi4orbUJvZHADHcUqAkbhN9ItmtjaZHTuO0w+4DjgC
   Nzay2HGc+Wa2spftLgOacRt2EYnhdbPjZ3MejUZ56qllwJvANoC3DVn32isqymltndLnjOneznXx
   YneedHs7/O53UFGRWR1h/wUnGck03V7O9BYRySvxxn8A/YHLgTbg6c5bGzAT6J9ofAhwENDc7fsL
   gQt72W4C8EPgVqA2zr68m4UikoO8Xj7bryWmY/cNpQYRX0flxS7v3df2AwbsY0cc8bp96Utmt9xi
   9vnn3tWRyvllezxdMvV6vdKjiEiuwsulx3GjGjcDA7vdtxVwE3B1wh3DCcBN3b4/Bbg2ZpvtgEdw
   r0zfCoyPsy8fnzaR8MuVJaZ72/eQISM8a7xSrX3T7dttp53usY8+ym4d3fn5C02i4/bVCAdVl4hI
   2KTTUPeVoT4G2NXMOrpdzf7IcZyzgOdxF3rp8+J3gp93Ne0Xmpk5juPQR+Rj+vTpG76urKyksrIy
   id2L5Idc/lP7fvvt42scoq3tXaqra4FNoxfu7+M7d9v6fnbZ5XYGDqz1rZ5EkolW+CFRXrtrRcmN
   MRY3UpRv0RYRkVgLFy5k4cKFme0kXqcN/Cudn3Xb5kB6Rj6mAlNitvk38HLnbQ2wGhjby778+AVE
   JKf0doUxnT+/ZzPy4fVVztj9O84gc5yBBpFNjvfPf5rtscd75jhLDf7q+SqNxcWDDA40ONCKiwcl
   vV+v/0IQ+x7o/n33OExf0Zi+9q2r1iJSaPA48nEfUNfL/acC8xPu2P3A40vATkAxsBTYo4/tFfkQ
   SUEmzY6fOVi/M7bNzc1WVlZhRUVDNzTSsLVBs8EcGz36DDvlFLNttzW79Vazv/7V+3rchnrYhue+
   uHhYIE1q7L6Ki4d1NvpzOp+brbrl2bfq9RePvvgZDxIRCSuvG+rhwBPAIuCKztsiYDEwPKmdwxjc
   eMiLwNTO+84EzuxlWzXUIiko5Gant3OHbxsssf79P7Jp08zWrMnu8VN57r36paP35+HAzq97+9n4
   lOot5PeYiBSudBrqvsbmve44zgHAYcBeuJno+83soRTiJA8AD8Tcd2OcbU9Ldr8ihawr0+qOphsb
   dDkhsTNwFf36PcbNN3/Id79bEXRBfcrW/OlM5XJ2X0Qkm/r6UCKdXfpDnTcRCVjPedQ7A+du+Fkh
   NTsbG71S3IVct2SXXX7BtdceR01Nhe8fpAtLoxlbh7vE+2esW9dE7PvD/foHbFzmPHG98T6oKCIi
   PcVdejxMtPS4iGvTpZ8nM2TIvey33z4FNYHhtdegru5NHn10c0aO/AMzZ+7GmDHuuSe7zHamwjL9
   IrYOoMfCN4sWPb3h67lzW3j11dfZccdtuPTSnxbM+0VEJBXpLD2uhlokh2zaUDdRVeX9kt5h9ckn
   cPnlcN118KMfwQUXwJZb9tym0J+jeLL1i4aISK5Lp6HuM/IhIuESlqhBtnV0wB/+AFOnwujRsGQJ
   7LBD0FXllqDmX4uIFIKioAsQkeR1ZVqrquZTVTW/IK4wPvYYHHQQNDR8yLbb/oR3361l5cpo3O0j
   kXpKSqYATWzMC9f7WmM0GqW6upbq6lqi0fi1iYhIflLkQ/KW3xnXsGRog+Tnc/Cf/8CFF8Lf/w4n
   nfQM115bw9q1vwISxxWy+drkSpQiV+oUEQlaOpGPlGbsBXVDc6glRdlesa8QV5Dz6zlYs8bsoovM
   hgwx+/nPzT7+ONzzkMNcWyy/F90REckHeDmHWiSX+Z0XVR7V++egowNuuw1+8hOoqIClS2H77T0s
   WHJm/rWISK5RQy0SAkHGR8IQXfnHP2DCBOjXD/74Rzcz3b2+trZ3KSqK0NGxHBgVqg9jFuoHRUVE
   pJtUL2kHcUORD0lRLkU+goyPZHJsL+p+5RWzb33LbPhws9tuM1u/vu9jFBUNtrKyQ0IXV1CUQkQk
   f5BG5EMfSpS8lSsfSgxybnKmx073Ofj4Y/jVr+D66+Hcc2HyZNhiC+/rExERSZXmUIsnwhAB8ILf
   eVHlUVN/Djo64He/g2nT4LDDYNkyGD7cxwJDKl/+jYmIiEsNtfQQO1qrtbVOo7V8FmQGN96x/Wj4
   /v53mDgR+veHP/0JDjig9+26H7uiopzW1ilxn5tcbEz1b0xEJP8o8iE96E/swQjThxIBT+cVv/wy
   TJkC//wnXHYZfPvb4MT5Q1pvs5KnTTuHRYue3lBfVx25OldZ/8ZERMJNkQ8pWLl4pbK7IOMjsceu
   rq71ZBzemjVw6aVw443uBI85c+ALX+j7Mb2N4lu0qPdms7dtp069NOdeexERyX1qqKWHXBwBpj+h
   h0tHh9s8X3QRVFXBM8/Adttl59jLlq0gGo2G+rXPxX9jIiKSQKpjQYK4obF5WZVrI8ByaaW6bMj0
   9ctkHN6iRWZlZWYHH2z2+OMpH7rbsSMGB1pR0VBraGiIu21R0eANdYL7uFx47XPt35iISCFBY/Ok
   ECmTupFXueJUIzT//jdccAEsXgyXXw4nnhg/J53IjBkz+NnPGunouDLhOZSXH8qSJeuBbYF64O2C
   fe1FRMQb6WSo1VBLzsvVD6f5Idu/XHz0Efzyl3DTTe4Ej0gESkoy22cq56DXXkREvJZOQ13kVzEi
   2VJTU8O8eW7TVVU1Xw1VEqLRKNXVtVRX1xKNRhPeH2v9erjlFthtN3j7bVi+3M1MdzXTye4nU7Gv
   /bRp59DYONuz42brPEREJMelmhEJ4oYy1CJJSSb/HG+bZLPTjzxitu++ZoccYrZ4cXo1ZHoO6Twu
   1dxykEvC+0G5bRGR5JBGhjrwZjmpItVQiyQtUeMU70OciT7c+eKLZuPHm+24o9ldd5l1dPR+fC8+
   JJpO89fXcdNpjvPpw6759suBiIif0mmoNTZPJM94PdP6o49gxgy4+WY3I33bbZnnpBPx+hx6m1md
   zmztXFXo5y8i4jc11CIp8HsBmWwsUNPXHOTu9w8YMJU992xht91gzBhYsQK+9KXM9u8nr4+redEi
   IpK0VC9pB3FDkQ8JAb//bJ7NP8vHi1R03b/ffj+znXf+yEaPNnvyycSPS3b/fuvrvNLNZedD7liR
   DxGR5KE51CL+8XskXRjmab/4Ipx/PixdCjNnQm3txnnSuT6iLteXp89UoZ+/iEiy0hmbp8iHiPDh
   h9DQALfeCpMnwx13wIABPbfJ9Ryu17nsXFPo5y8i4ifNoRZJUiRST0nJFKAJaOrM1NantI++5hp7
   sf9UrV8PN97ozpNevvx19trrdB5+uJZFi9Kbudx1fuXllYwcuRdDh46kvLySGTNmZHWes9/zozWf
   WkREekg1IxLEDWWoJSQyydQmOyM6W5ndBx80GzXK7OtfN7vuun+kPb+6959HDLbasK37dSQr+d18
   yrqLiEj2oTnUIuEVlrnG//qX2dixZjvvbDZ3rjtPOtna+mr4e+5j0/259/l/3n4/z2F5HUVExB/p
   NNTKUIvkgWQ+cPbBB3DJJdDU5H7w8K67Ns1JJ6IcroiIyKaUoZaClW4ONt3H+ZWR7pq+0dIylpaW
   sYwbV9ejrs8/h+uvd3PSH33kzpOeMqVnMx1bW3Hx+bS1rU7pHHvuY2fg3A37c7/e2dPzTq4O748X
   u/+ioolUVJT3um00GqW8vLIzS36o8tYiIvkq1UvaQdxQ5EM8lslc4kzys35kpPuKICxYYPaVr5hV
   VpotWZJcbWVlFVZcPCitc+y+jxEj9rQhQ0ZYWVmFNTQ0ZHWes99Z9IaGBisqGmpwoEEkbua8uHhY
   txx5qRUXD1LeWkQk5NAcapHkpDvzOQyzopOp6eCDFzN06HU89xzMmgXHHbdxnnQ6+wv6HMMmmeeo
   t23gBqqqttVzKSISYunMoVbkQyTHVVSUU1QUAQ4Cfkq/fu2sWHEFo0fDs8/C8ccn30yLiIhIGlK9
   pB3EDUU+Co7ff7LPVuQje+fRZPCYwWr76lefsrffTu6xmS7TnS9Lc6cq2RGIinyIiOQeNDZP8kG2
   5vym2wwm+7hszEMeMmSEwbkGazrz0/clNcItmZnSic6x0OcxJ/sclZVVdGbJDymo50dEJFel01Ar
   Qy2hky8ZXj/PIxqNctxxl/Dpp3cAXwTOAU4G3s5aFjxfXicREZHulKGW0NOSzZl77z2or/+MTz99
   ENge2Ao4DJie5oi4KHADTz21LG9fE73vRETEV6le0g7ihiIfeSHZiEC+RAm8Po9168yuvdZs2DCz
   4cObDe7sMSpvyJARCffffazdZpttYbC7QfIj8rrHHBoaGnLidcqX95OIiGQHylBLmKWyZHO+fNjN
   q/N44AGzPfYwO/xws2eeSa9JjH0MbGXwlZRek9hjZnu+dDq0VLiIiKQinYZaS49LKOXLEteZnsfK
   lRCJwIsvQmMjHHOMOwJv1Kga5s1r6rbceFPC4zQ2zqa9/TI2Zp4BLkm6ltjHt7fDokXKTIuIiKih
   lqyJROppba2jvd393l3eeleqq2uJROrzooH2yh//+DCRSDtvv30op532FvfeuzvFxT238eaXjm2A
   KRu+czPYTRnuM1xi33f5eI4iIhIsTfmQrIpGozQ2zqat7V2efXYZ69ZdBbhNzrx5ia+y+lEL0GtD
   n+jnftQzc+YtrFp1KG+8cRLwN+Bmiov/yfz5d2Z8/Gg0yrhxdZ1XmQHOBbYF9qKoaCH77PMVLr10
   6objxJ4/0OPxQbxm6cr2a5lL9NyIiPSUzpSPwPPRydxQhjrvBJ1rTWYOczY/yOYuAnKSwfsGqwwO
   6LEgSFnZIZ4dp6yswoqKhhpEDOZYUdFga2ho2GS73s4/X7Lt4tIHNkVENoUy1CLJ6S0P3Ng4e8PV
   uUQ/99Jzz8F3vvN/rFt3Pe5M6dOBs+medX711eSzzn2pqamhsXE2HR2NG/bf0eFmoadN27hdvPNf
   sGCurmDmkWy+z0VE8pnmUEsgIpF6SkqmAE1AU5rzk72S/hzmvuYbJ5p93NYGP/4xVFTA0KFLgPlx
   j7PjjsNTqktERESyKNVL2kHcUOQjLwUZH9j4p+6IQekmf/JuaGjoHCu3ccRcsrGIRD9bt87sqqvM
   SrBxaYIAABlvSURBVEvNfvQjs3feid0+0uPYxcXDkp4NnczzmMyf+b1YmlzCT5EPEZFNoTnUIslr
   bm62IUNG9JrldjPeEYPxnbfIJhnvvnLgvf3siCPG21/+YrbbbmbV1WYrVmxaT/dFU5JpWNNtiJJp
   iONtoyYsv+iXIxGRntJpqDXlQwpadXUtLS1j2ZhXbqKqyo1exN5fVnYTpaVbA25kpbFxdq+PXbBg
   bi/7vY+hQ3ektHRfrrgCxoxx50n7Vb+fs6GDOKaIiEi2pDPlQx9KlLyTyhiwvmYU95yZPYFnn+3P
   unU/2PCzadPOobV1Sq+P3bjfzYF9gUP45jff4ZproH9/f8/JD92P39a2OqvHFhERCb1UL2kHcUOR
   D0lSuktyx4s2dN1fVlbRa7wj3mM//dSsvn6l9e//gW2//V/s7rsf8uWcshG/iD1GcfEwKy4epMiH
   iIjkJRT5kELnVxwh2f2awV/+4i4XPnKku1z4HntkdOiEx/b76nVvxy8ru5XS0qG+HVNERCQoinxI
   VgQdPwhCMstXL18OkybBG2/A1Ve7Oels8GYJ8tSUlg7t85eUQnyPiIhIAUv1knYQNxT5CI2wT3hw
   VxwcZHCgwYFWXDwoo/qSmbzx3/+anXWW2bBhZtde647F81KyY+78mtSQ6muezfdIpuetCRciIhIL
   jc0TvwW9ZHgibkM9LOkZzon21Vdj+OmnZrNmufOkzzvP7N1306853RF2ydTphVQaz2y9RzI977D/
   cigiIsFQQy2+C3tD7WV98fbV0WF2771mI0eaHX202cqV6dfrRVMXttckW/VkepywPW8iIhIO6TTU
   ylBLSpLJEuezNWt25Igj4O234brrINNocGPjbNrbL6PrA3/t7e59uZw5LvT3iIiIFB411JKSmpoa
   5s1r6vaBs6ZQNX9eNnMVFeU89NBEOjoABtCvn8PKlZcyYwaceSZsFpJ/PWFrYLP1Hsn0vMP2vImI
   SO7S2DzJO6lMmIi3bTQaZdy4Otrb64DtgZM4+OAX+ctfDmDwYG9rdY9zGeA2dfPmpd6AFupUjUzP
   u1CfNxERiS+dsXlqqKVg9dXMVlXV8uCDPwIO69x6HlVVt/myvLaaOhERkfDQHGqRFMTLL2+zTQ1P
   PTUdGN5t64+S2mc6zXEQc6RFRETEO0VBFyASHgN47rmzqK6GQw99HcfZFWgCmjrztfV9PrrrindL
   y1haWsYyblwd0Wg0K5WLiIhIcHSFWgrWxg+lFQF7AEdwwAFrOPnkhzj11NMwqwNuoKjoBaZNm5jw
   KnI+TuwQERGRxHSFWgpWdXUNkcgDlJRUMWxYBzff/Bxz5+7EjTf+prMxngU8RkdHI4sWPR10uSIi
   IhJSaqilIC1ZAt/4Btx7bxnz52/Df//7NU4/fXRG+4xE6ikpmUIqMZFcFo1Gqa6upbq6VtEWCRW9
   N0Uk29RQSyj59T/Et9+GM86AMWPgpJPcxvqII3puk25j3DV/uapqPlVV89Maf5crwpIXV+MkscLy
   3hSRwqKxeRI6Xs1m7m7tWrjqKpg1C047DaZNg0GD+q5Bo+ziq66upaVlLF15cXB/kfBjrGA8frxP
   JPeF4b0pIrktlGPzHMc5ErgK6AfcbGaXxfz8O8AFgAOsAc42s2f8rkvCy8sP95nB3LlwwQWwzz7w
   z3/CyJGJH6dRduGnD4GKiEhY+NpQO47TD7gOOAJ4A1jsOM58M1vZbbN/A183sw87m+/ZwIF+1iWF
   4emnYcIE+PBDuPlmOOywxI+R5GjZbgkrvTdFJAh+Z6i/BrxoZq+Y2WfAncBx3Tcws8fM7MPObx+n
   52oakgVhy6Fm+uG+t96C738fjjoKTj3VbazVTHsrDHnxQvsQqCQnDO9NESk8vmaoHcc5Aagxsx90
   fn8KcICZnRNn+8nArmZWH3O/MtQ+CWsONZ0M89q1cMUV0NjofvDwJz+BL37R70olSMq6i4iI19LJ
   UPvdUNcCRybTUDuO8w3g18AhZvZ+zM/UUPskHz7AYwb33OPmpMvL4fLLYcSIoKsSERGRXBTGDyW+
   AWzf7fvtgddjN3IcZ2/gJtzm+/3YnwNMnz59w9eVlZVUVlZ6WafkgN6uRj75JEycCGvWwK23gt4W
   IiIikoqFCxeycOHCjPbh9xXqzYDngcOBN4EngJO6fyjRcZwdgIeBU8zsn3H2oyvUPglr5CNWbJ0D
   BlzB6NF/Zfny7bjkEncUXr9+ifeheICIiIj0JXSRDwDHccawcWzeLWZ2qeM4ZwKY2Y2O49wMjAP+
   0/mQz8zsazH7UEPto1xoNDeNpqxlp50eYNmycWy1VeJzyJVfHCT/5cK/NxGRQhbKhtoLaqilqqqW
   Bx88F6jovGcuVVW3s2DB3KSa5U0b8skMGXIv++23j5oayRr9YiciEn7pNNRaelx80dsovnTH8y1e
   DG++eTOOMxhoxh2R9qMNI9J6LvDhNitdVwDjVAc08d57P427NLFXowST2U/YxhaKf1J/r4qISC7w
   faVEKTyxV+FaW+uYNu0cZsy4tsd9ia7MvfGGO/qupQUaGgazzTZPcNVVNwH8f3v3Hm11WSZw/PtI
   aKhdNFZOFzMnC0RNUdM0HXXV4YglBWRp5ThTa5xUUgsQmXEy7zeIUjPUUnRWYcvShlwGMiihJDYq
   l6OIibdkbExCU8kUOO/88fsdORzPZe9z9n1/P2udtfblt3+/d2/edfZzHp73eZk4sbis3pabPcwE
   ptHTDnvdjb8/WcRCzlOqa0mSpCpKKdX8TzZM1YuWlnEJZqWsoV1KMCvtuOOH3vRYS8u4bl+/fn1K
   556b0o47pjR1akovvdT79ebOnZuGDNkpP/+sNGTITmnu3LndHtfSMq7PsXQ3/p7GWuzn0PU8pbqW
   6kOhc1WSVD153FlUrGqGWjUjJbjpJjjzTDjwQLj/fth1175f17Ez2vTp17B27XPA8Df+G71zpre1
   tZXW1tZOWeHscbcmVqV0nqtQ/P+0SJJqVLEReDV+aOIMdUdWtaVlXN1ksrrLwp1//vm9ZuaWLEnp
   oINS2m23F9P++/97v95vMdm/3j7XUmURCzmPGUtJkmoL/chQ2+WjhtVzR4DuWoN199iaNTB1Ktx5
   Jxx7bBtXXTWKv/3tYqD499vdro8jR17P0KHv2uKa/R1/fxRyHtuoSZJUO2yb12AaYVvwnvz1r3DZ
   ZXD55XDSSTBlCowfP7D3293ntdVWE2lvnw7U1x8kkiSpOmpx63FpC+3tMHt2lpU++GB44AH44AdL
   c+4tO3nAVlt9k/b2r9JTNw9JkqRSMKCuYV0DxHpfPLdkCZx+OmzaBD/9KRxyyJbP9/R+Cy2J6Lrg
   a+3aESxdulfZ3k+tsGREkqTqsuSjxjVCsPTMM1nnjoUL4cIL4fjjYasethTq+n6BfteR13MNeqGa
   4T1KklRJ1lCrpqxfn9VJX3EFnHxyVie9/fbFnWOgdeSN8AdJbxq5zl6SpGqwhlo1ob09K+mYOjUr
   63jwQdhll+qMpaP3tCRJUrkYUKuk7r03q5NOCX72s2zh4UA0Wh15qfn5SJJUfZZ8qCT+8IesTnrR
   IrjoIvjyl3uuky5Wo5dtDJSfjyRJpWMNtSrulVfg0kvhBz+ACRPgjDNgu+2qPSpJkqT+6U9AXaIc
   oppNezvceCMMHw6PPw5Ll8I55xhMS5Kk5mMNtYq2eHFWJz1oENx8Mxx00ObnLD+QJEnNxpIPFezp
   p7PWd4sXw8UXw3HHbVknbU9kSZJU7yz5UFm88gqcdRbsuy/svjusWtX9osPp06/Jg+kTgCyw7shW
   S5IkNSoDavWovR1mzYJhw7Ls9PLlcPbZ1klLkiR1Zg21unX33fDNb8LgwXDLLXDggX2/xp7IkiSp
   GVlDrS089VTW+m7JErjkEjj2WIgiqohclChJkuqZfajVby+/nC00nDkTTjsNJk2Cbbet9qhUTv7x
   I0nSm7koUUVrb4frr8/qpNesgRUr4NvfLk8wPW/ePEaNGs+oUeOZN29e6S+ggnV0ZJk/fwzz549h
   7NgT/DeRJKmfzFA3sUWLsn7SQ4bAjBlwwAHlu5Yt9WrLqFHjmT9/DFlHFoAbaGmZwx13/KKaw5Ik
   qerMUKsgTz4JxxwDxx+f1Uvfc095g2mwpV5vzNxLklTf7PLRRF56CS66CK65JuvgceONWXZa1dM1
   c3/PPSdUJHNvRxZJkkrHDHUT2LQJfvxjGD4c/vhHaGvLNmoZaDBdTGZ14sQTGTJkCnADcEMewJ04
   sAE0gGpl7ltbW7n11qzMo6VljuU3kiQNgBnqBrdwYZaN3m47mDMH9t+/NOctNrPaEcBt7iqx+Vi7
   TVRHa2urn7UkSSXgosQG9cQTMHkyPPAAXHppVjNdTD/pvpRqUVuzL1Zs9vcvSVKtcVGieOklmDIF
   PvYx2G8/eOQR+MIXShtMv9k8YCYPPLC86EV1zb5Y0dILSZLqnyUfDWLTJrjuuqyH9OjR8NBD8J73
   lO96hx22LwsWTKS9fRrwB+By1q2DsWMrs6iukVh6IUlSfTOgbgB33ZX1k37HO+C227LMdDnNmzeP
   Cy64gvb26fkjk4C/A1p59dUs61xogGi3CUmSVO8MqOvY6tVZnfSyZXDZZTB+fLlLOzJblml0uAYo
   Psva22JFSZKkemBAXYf+8he44IKsxGPSJJg9G9761i2PqXznjGfZ3A6vuAyzJQ+SJKmeuSixjmza
   BFdfDcOGwbp1WZ30mWd2H0yPHXsC8+ePYf78MYwde0JJd+Dr2lN6660nM3LkoF4X1bkboCRJalS2
   zasTCxZk/aR32AG+9z0YObLnY0vV0q6zrhlvoOAMuK3hJElSvehP2zxLPmrcY49lZR1tbVmd9Lhx
   lamT7qynTVwKDdC71lwXu3BRkiSpllnyUaNefDELpA86CA4+GFauLHzRYam3+W72XtGSJEm9MaCu
   MRs3wsyZMHx4tvjwoYeyjVq61kn3ptY2Cyl1gF/PrCWXJKnxWENdQ+bPh299C4YOhRkzYJ99qj2i
   TClqoCvfdaT2WEsuSVLt608NtQF1Dfj977PyjocfhmnT4HOfq3yddF/KGRA3S7BdjsWikiSptFyU
   WGdeeAHOOw9uvDEr67j5Zthmm2qPqnvl6hXd04LHRg2qJUlS47GGugo2boSrrsrqpNevzzLTkyfX
   bjBdTs204NFackmSGpMZ6gq7446sn/ROO2W399672iNSpbjNuiRJjcka6gp59FGYOBFWrYLp02HM
   mNqrk64GF+pJkqRa4qLEGrRuHZx7LvzkJ9k24RMmNGdpR2+aZVGiJEmqfQbUNWTDBrj66iyYHj8e
   zjkH3v3uao9KkiRJvbHLR42YOzfrJ/3e98KCBbDXXtUekSRJksrFgLqEVq3KAunVq7M66c98xjpp
   SZKkRmfbvBJYtw5OOw0OPRQ+9alsu/CjjzaYliRJagYG1AOwYQNccUXWT3rjRli5MstQb711tUcm
   SZKkSrHko59+/esseN55Z7jzTthzz2qPSJIkSdVgQF2klSuzftJPPJHVSX/605Z2SJIkNTNLPgr0
   5z/DN74Bhx0Gra3Q1uaiQ0mSJBlQ92nDBvj+92H33SEleOQROP1066QlSZKUseSjBynB7bdn5R27
   7AJ33QV77FHtUUmSJKnWGFB34+GHswWHTz8N3/0ujB5taYckSZK6Z8lHJ2vXwimnwBFHZIsN29rg
   qKMMpiVJktQzA2rg9ddhxoysTnrQoKxO+tRTYfDgao9MkiRJta6pSz5Sgttuy+qkd9sNFi3KgmpJ
   kiSpUE0bULe1ZXXSa9ZkXTxGj672iCRJklSPmq7k4/nn4eST4ZOfhM9+FlasMJiWJElS/zVNQP36
   69nOhiNGZD2kV62CCROsk5YkSdLANHzJR0owZw5MmgTDhsHdd8Pw4dUelSRJkhpFQwfUK1ZkddLP
   PgtXXpltGS5JkiSVUkOWfPzpT/D1r0NLC4wdmwXWBtOSJEkqh4YKqF97DaZNy+qkhwzJ6qRPOQXe
   0tB5eEmSJFVTWQPqiDgyIlZFxGMRMaWHYy7Pn18eESP7c52U4Je/hD32gN/8BhYvzjZq2WGHgY1f
   kiRJ6kvZAuqIGARcCRwJjACOi4jduxxzFLBbSunDwInAD4u9zvLlWQu8s86Cq66CX/0qW3yoxrdw
   4cJqD0E1yHmh7jgv1B3nhUqlnBnqA4DVKaWnUkobgJuAz3Y5ZgxwA0BK6T7gnRGxUyEnf+45OPFE
   GDUKjjkGli3Lbqt5+ItQ3XFeqDvOC3XHeaFSKWdA/T7gmU731+SP9XXM+3s76WuvwaWXZuUdb3sb
   PPoonHSSddKSJEmqjnKGoanA46KQ16UEt94KkyfDnnvCb38LH/nIwAYoSZIkDVSkVGjcW+SJIz4O
   fCeldGR+fyrQnlK6pNMxM4GFKaWb8vurgMNSSs91OVd5BilJkiR1kVLqmvDtVTkz1PcDH46IDwLP
   Al8EjutyzBxgAnBTHoC/2DWYhuLflCRJklQpZQuoU0obI2ICMA8YBPw4pfRIRPxr/vzVKaXbI+Ko
   iFgNrAf+uVzjkSRJksqhbCUfkiRJUjOoqZ0SK7URjOpLX/MiIr6cz4cVEbE4Ij5ajXGqcgr5XZEf
   97GI2BgR4yo5PlVHgd8hh0fE0oh4KCIWVniIqoICvkOGRsTciFiWz4t/qsIwVUERcV1EPBcRbb0c
   U1S8WTMBdaU2glF9KWReAE8A/5BS+ihwHnBNZUepSipwTnQcdwkwlzd3E1KDKfA75J3AD4CjU0p7
   Ap+v+EBVUQX+vpgALE0p7QMcDkyPCJvxNrbryeZEt/oTb9ZMQE2ZN4JR3epzXqSU7k0p/SW/ex99
   9DJX3SvkdwXAN4CfA89XcnCqmkLmxZeAX6SU1gCklNZWeIyqvELmxR+Bt+e33w78OaW0sYJjVIWl
   lO4GXujlkKLjzVoKqMuyEYzqXiHzorOvAbeXdUSqtj7nRES8j+xLsyOr4GKRxlfI74oPAztGxF0R
   cX9EHF+x0alaCpkX1wJ7RMSzwHLgtAqNTbWr6Hizlv5Lo6QbwahhFPzvGxFHAF8FPlG+4agGFDIn
   vgecmVJKERFY8tEMCpkXg4F9gU8C2wL3RsSSlNJjZR2ZqqmQefFvwLKU0uER8SFgfkTsnVJ6ucxj
   U20rKt6spYD6f4GdO93fmewvgt6OeX/+mBpXIfOCfCHitcCRKaXe/htH9a+QObEfWX97gKHA6IjY
   kFKaU5khqgoKmRfPAGtTSq8Cr0bEImBvwIC6cRUyLw4GLgBIKT0eEU8Cw8j201BzKjrerKWSjzc2
   gomIrck2gun65TcH+Ed4YyfGbjeCUUPpc15ExAeAW4CvpJRWV2GMqqw+50RK6e9TSrumlHYlq6M+
   yWC64RXyHfJfwCERMSgitgUOBFZWeJyqrELmxSrgUwB5newwssXual5Fx5s1k6F2Ixh1p5B5AXwb
   2AH4YZ6R3JBSOqBaY1Z5FTgn1GQK/A5ZFRFzgRVAO3BtSsmAuoEV+PviQuD6iFhOlmg8I6W0rmqD
   VtlFxGzgMGBoRDwDnE1WEtbveNONXSRJkqQBqKWSD0mSJKnuGFBLkiRJA2BALUmSJA2AAbUkSZI0
   AAbUkiRJ0gAYUEuSJEkDYEAtSWUUEZsiYmn+82BE7BIRi/PndomI4zodu3dEjO7HNRZGxH5dHjs7
   Ii7s8tg+EdFj3+WI+E5ETCz2+pLU7AyoJam8/ppSGpn/7JtSejql9In8uV2BL3U6diRwVD+ukfKf
   zn5KtitcZ8fmj/d2HklSkQyoJanCIuKV/ObFwKF59voM4Bzgi/n9YyJiu4i4LiLuy7PbY/LXD4mI
   myJiZUTcAgwBovM1UkqPAS9EROddQ48BZkfEv0TE7yJiWUT8PCKGdH5pfo03st4RMTQinsxvD4qI
   y/LXL4+IE/PH3xMRi/Kxt0XEIaX+3CSpVtXM1uOS1KCGRMTS/PYTKaXxbM4ETwEmpZSOBoiI54D9
   Ukqn5vcvBBaklL4aEe8E7ouI/wa+DrySUhoREXsBD9J9dnk2WVb6dxHxcWBdSunxiHghpXRtfo3z
   gK8BV3Z5bXdZb/JjX0wpHRAR2wD3RMQdwDhgbkrpwogIYLsiPydJqlsG1JJUXq+mlEb28Fx0c7/z
   Y6OAoyNiUn5/G+ADwKHA9wFSSm0RsaKH8/8M+G1eF9253GOviDgfeAewPTC3iPczKn/95/P7bwd2
   A/4HuC4iBgO/TCktL+KcklTXDKglqXZ0lxEel5dvvCFLAL8pGH/zyVJak5dqHE6WQf54/tQsYEwe
   jJ+QP9/VRjaXBb61y3MTUkrzu74gIg4FPgPMiojvppT+s68xSlIjsIZakqrnZeBtvdyfB5zacSci
   OjLdi8gXM0bEnsBHe7nGbGAG8HhK6dn8se2B/8uzyV9hcyDfOUP+FLB/frsjG90xppMj4i359T8S
   EdtGxAeA51NKPwJ+RLbAUpKaggG1JJVXd1nnjseWA5vyxYGnAXcBIzoWJQLnAYMjYkVEPES2aBHg
   h8D2eQu8c4D7e7n+z4ERZIF1h/8A7gPuAR7pMq6OsU0DToqIB4F3dXr8R8BK4MGIaMvH8hayLPey
   /PgvkJekSFIziJTskiRJkiT1lxlqSZIkaQAMqCVJkqQBMKCWJEmSBsCAWpIkSRoAA2pJkiRpAAyo
   JUmSpAEwoJYkSZIGwIBakiRJGoD/Bw7Ms9t2AKdoAAAAAElFTkSuQmCC
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h4 id="Plot-fitted-values-vs-Pearson-residuals">Plot fitted values vs Pearson residuals<a class="anchor-link" href="#Plot-fitted-values-vs-Pearson-residuals">&#182;</a></h4>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>Pearson residuals are defined to be </p>
   <p>$$\frac{(y - \mu)}{\sqrt{(var(\mu))}}$$</p>
   <p>where var is typically determined by the family. E.g., binomial variance is $np(1 - p)$</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[40]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">,</span> <span class="n">title</span><span class="o">=</span><span class="s">&#39;Residual Dependence Plot&#39;</span><span class="p">,</span> <span class="n">xlabel</span><span class="o">=</span><span class="s">&#39;Fitted Values&#39;</span><span class="p">,</span>
                             <span class="n">ylabel</span><span class="o">=</span><span class="s">&#39;Pearson Residuals&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">scatter</span><span class="p">(</span><span class="n">yhat</span><span class="p">,</span> <span class="n">stats</span><span class="o">.</span><span class="n">zscore</span><span class="p">(</span><span class="n">glm_mod</span><span class="o">.</span><span class="n">resid_pearson</span><span class="p">))</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">axis</span><span class="p">(</span><span class="s">&#39;tight&#39;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">([</span><span class="mf">0.0</span><span class="p">,</span> <span class="mf">1.0</span><span class="p">],[</span><span class="mf">0.0</span><span class="p">,</span> <span class="mf">0.0</span><span class="p">],</span> <span class="s">&#39;k-&#39;</span><span class="p">);</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAswAAAG8CAYAAADKG/NGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzt3XmcHGd95/Hvb2wLD4cPWQ6nscGQGIMRA4ljIkBDkhkZ
   AiLyQLgjIIs2ZBcINEQYlqAsApuAOJa8FmJD8MQhkCzGILJkWsMxBkHiAJYPbGC5j3AKA+EYItv6
   7R9VLbVa3dXV3XU8VfV5v171mumznnrq+vVTv3oec3cBAAAA6G+q7AIAAAAAISNgBgAAABIQMAMA
   AAAJCJgBAACABATMAAAAQAICZgAAACABATOARjGzp5pZO+H1FTP7owzmM2tm35z0e0JgZjvM7PKy
   y5FGneodQDgImAEEy8y+Zma/MLOfmtl3zexyMzthku9093e6+6akt8RTrszsoJn9LF62/Wb2ITP7
   g7znO6agOuw3s2eY2W1x3f3EzPaZ2e+N8T2Xmdkr8ygjgHohYAYQMpf0GHe/k6T1ks6R9D/KLVKm
   Hhgv269KukzSX5nZn5dbpMr4RFx3J0l6u6R/NLOTSi4TgJoiYAZQCe7+PUl7JN2/85yZnWdmnzSz
   H5nZtWa2seu1Z5jZl83sP8zsK2b2lK7nP971vjkz+7yZ/djM3izJul47IhXBzM6IW4an4sfPNLOb
   4nl82cy2jblsN7v730l6jqQLzWxt/P0nmtnbzezbZvYtM3tl17yfYWafMLM3x2X/nJn9dldZh312
   r5m91sxujuvn/K7P3svMroqXa4+kdd3lHVLvK2b2P+Pv/w8za5vZKV2vP6zrs98ws63x87czs9eZ
   2dfjqwlvMbPjE6rN4rpzSe+QNC3p3ke9yex+cZl+ZGafNbPHxs9vk/QUSX8Wt1S/P9XKAtBIBMwA
   QmeSZGb3kHS+pKvjx3eX9E+S/qe7nyzpRZKuMLNTzOwOkt4k6Xx3P0HSQyVde9QXm62TdIWkl0o6
   RdKXJW3oesuwVITvSfq9eB7PlPQGM5sZd0El7ZZ0rKTfiB9fJumApDMlzUial/Rfut5/rqQvxWV/
   haT3drWypvns5+PP/qWiVtqOv5f0qfi1V0raqrgukuq96/NPlvQMSb8iaU38HpnZ6ZI+qGjdrJP0
   IB1eLxdLuo+iKwn3kXR3SUNb283s2Hi5firpiz2vHSfpA5KWJJ0q6bmS3mlmv+rul0h6p6TXuPud
   3P1xw+YFoLkImAGEzCS9z8z+Q9I3FAW0O+PXnibpg+6+JEnu/iFJn5b0e4qCu4OSzjGzaXf/nrvf
   1Of7Hy3ps+7+Xne/zd3fKOm7PfMfyN0/6O5fjf//mKIW8IePuaxy91sk7Ze01szuLOlRkl7g7qvu
   /gNJb5T0pK6PfN/d3xSX/R8lfUHSY1J+9uvu/va4hfZvJd3VzH7FzO4p6dclvdzdb3H3jysKOjuS
   6l2K6v4d7v4ld/+lpH9UFBhLUYvusrv/Q1zmm939OjMzSc+W9EJ3/7G7/0zSRT3l7XWemf1I0nck
   PVHSFnf/ae97JN3B3S9291vd/aOKgv0nx6+bhqxjAJCilgwACJVLepy7f8TMHqEocPt1Sf8m6XRJ
   T+hcYo8dK+kj7v4LM3uiopbNt5vZJyS13P0LPd9/N0nf6nkudQ8LZvYoRS2791XUAHF7SdenXrqj
   v+84RS2hNytavuMkfSeKJ6V4Ht/o+si/93zF1xUt0z1TfPbQD4O4viTpjopahX/k7qs933ta/P/A
   eu/33ZJW4+9V/B1f0dFOVVR3n+kqrym5Uedf3X3Yj5O76ej12akjKbCbGQGEi4AZQCW4+8fiHOPX
   SHqkouDvcnfvmzfs7nsk7TGz20l6laRLJT2i523flnToUnzc0nla1+s/UxTIddyl6723U5TO8TRJ
   73f328zsSk3WYvk4Sbcq+kFwvKT/lHSKux8c8P679zw+XdL7FQWJwz47yHcknWxmt3f3X3R9723x
   /4n1PsQ3FKWC9NqvKLA+292/M8b3DvJtSaeZmcUt6VK0LJ+P/ydgBpAKKRkAquSNks41s9+U9HeS
   Hmtm82Z2jJkdb1EfvHePUwseF+cy3yLp5zoc8HX7oKT7m9mWOBf2eeoKihXl1z7CzE4zsxMlXdj1
   2pp42i/pYNzaPD/i8nTys9ea2VMl/ZWki939R3HguEfS683sTmY2ZWZnxi3tHb9iZs8zs+PM7AmS
   zlKULvHdFJ/ty92/rijF4i/i732YpMd0vWVgvfcuVx9/L+l3zewJZnZsnG++Pg7qL5X0RjM7Na6T
   u5vZqPXZ62pJv1B0Y99xZjYbL8u749e/pz43CgJALwJmAJXh7vslLUra7u7fUtQi+1JJ31fUetnS
   4Uv5L1CUsvBDRXnFz+l8TTx1vu8Jim4426/oZrO9XfP7kKR/UJRm8SlFKSGdz/5UUYD9j4pSKJ6s
   qHX3iCIPWaTrzKxzs9qzJP2pu+/oev0PFQXlN8Xz+D86MqC/WlE6yA8U3Zy34O4/SvHZfn1Ndz9+
   iqTfjD/354rqXPFyJ9V7v+/qru9vKMobbylaL/skPTB+33ZFNzD+q5n9RNKyou72+hnWV3Znfgck
   PVZRPvcPFP0gebq7/7/4fW+XdHbcg8Z7E74PQMPZ4atUAICqMLNnSPqjFHm8AIAJ0cIMAAAAJCBg
   BoBqKmQIbwAAKRkAAABAIlqYAQAAgARB98NsZjR/AwAAoBDu3rdbzOBbmN09iOkVr3hF6WWo4kS9
   UXfUXXUm6o26o96qM1F32ddbkuADZgAAAKBMBMwAAABAAgLmlGZnZ8suQiVRb+Oj7sZH3Y2Hehsf
   dTce6m181N14xq23oLuVMzMPuXwAAACoBzOTV/WmPwAAAKBMBMwAAABAAgJmAAAAIAEBMwAAAJCA
   gBkAAABIQMAMAAAAJCBgBgAAABIQMAMAAAAJCJgBAACABATMQM20223Nzy9ofn5B7Xa77OIAAFB5
   DI0N1Ei73daWLVu1uvoaSdL09HZdeeWiNm3aVHLJAAAIW9LQ2ATMQI3Mzy9oeXmzpK3xM4uam9ut
   PXuuKLNYAAAELylgJiUDAAAASHBs2QUAkJ1Wa5v27t2q1dXo8fT0drVai+UWCgCAiiMlA6iZdrut
   XbsukRQF0OQvAwAwHDnMAAAAQAJymAEAAIAxETADAAAACQiYAQAAgAQEzAAAAEACAmYAAAAgAQEz
   AAAAkICAGQAAAEhAwAwAAAAkIGAGAAAAEpQeMJvZMWa2z8w+UHZZAAAAgF6lB8ySni/pJkmMgQ0A
   AIDglBowm9k9JD1a0tsk9R27GwAAAChT2S3Mb5D0YkkHSy4HAAAA0FdpAbOZPUbS9919n2hdBgAA
   QKCOLXHevyVps5k9WtLxkk4ws7919z/sftOOHTsO/T87O6vZ2dkiywgAAIAaWllZ0crKSqr3mnv5
   99qZ2UZJL3L3x/Y87yGUDwAAAPVmZnL3vlkPZecwdyMyBgAAQHCCaGEehBZmAAAAFKEqLcwAAABA
   cAiYAQAAgAQEzAAAAEACAmYAAAAgAQEzAAAAkICAGQAAAEhAwAwAAAAkIGAGAAAAEhAwAwAAAAkI
   mAEAAIAEBMwAAABAAgJmAAAAIAEBMwAAAJCAgBkAAABIQMAMAAAAJCBgBgAAABIQMAMAAAAJCJgB
   AACABATMAAAAQAICZgAAACABATMAAACQgIAZAAAASEDADAAAACQgYAYAAAASEDADAAAACQiYAQAA
   gAQEzAAAAEACAmYAAAAgAQEzAAAAkICAGQAANFK73db8/ILm5xfUbrfLLg4CZu5edhkGMjMPuXwA
   AKCa2u22tmzZqtXV10iSpqe368orF7Vp06aSS4aymJnc3fq+FnJASsAMAADyMD+/oOXlzZK2xs8s
   am5ut/bsuaLMYqFESQEzKRkAAABAgmPLLgAAAEDRWq1t2rt3q1ZXo8fT09vVai2WWygEi5QMAADQ
   SO12W7t2XSIpCqDJX242cpgBAACABOQwAwAAAGMiYMbY6L8SAAA0ASkZGAv9VwIAgDohhxmZo/9K
   AABQJ+QwAwAAAGOiH2aMhf4rAQBAU5CSgbHRfyUAAKiLIHOYzex4SVdJup2kNZLe7+4X9ryHgBkA
   AAC5SwqYS0vJcPdfmtkj3f0XZnaspL1m9jB331tWmQAAAIBepd705+6/iP9dI+kYSTeXWBwAAADg
   KKUGzGY2ZWbXSvqepI+6+01llgcAAADoVWovGe5+UNKDzOxESW0zm3X3le737Nix49D/s7Ozmp2d
   LbKIAAAAqKGVlRWtrKykem8wvWSY2cslrbr767qe46Y/AAAA5C7IgUvMbJ2ZnRT/Py1pTtK+ssoD
   AAAA9FNmSsZdJS2a2ZSiwP1yd/9wieUBAAAAjhJMSkY/pGQAAACgCEGmZABAk7Tbbc3PL2h+fkHt
   drvs4gAARkALMwDkrN1ua8uWrVpdfY0kaXp6u668cpHh5AEgIEEOjZ0GATOAOpifX9Dy8mZJW+Nn
   FjU3t1t79lxRZrEAAF1IyQAAAADGVOrAJQDQBK3WNu3du1Wrq9Hj6entarUWyy0UACA1UjIAoADt
   dlu7dl0iKQqgyV8GgLCQwwwAAAAkIIe5wuiKCgAAoFy0MAeMrqgAAACKQUpGRdEVFQAAQDFIyQAA
   AADGRLdyAaMrKgAAgPKRkhE4uqICAADIHznMAACgEDT0oKoImAEAQO7o3QlVRsAMAAByR+9OqDJ6
   yQAAAADGRC8ZAAAgE/TuhLoiJQMAAGSGm/5QVeQwAwAAAAnIYQYAAADGRMAMAAAAJCBgBgAAABIQ
   MAMAAAAJCJgBAACABATMAArRbrc1P7+g+fkFtdvtsosDAEBqdCsHIHftdltbtmzV6uprJEWDGVx5
   5SL9swIAgkE/zABKNT+/oOXlzZK2xs8sam5ut/bsuaLMYgEAcAj9MAMAAABjOrbsAgCov1Zrm/bu
   3arV1ejx9PR2tVqL5RYKAICUSMkAUIh2u61duy6RFAXQ5C8DAEJCDjMAAACQgBxmAAAAYEwEzAAA
   AEACAmYAAAAgAQEzAAAAkICAGQAAAEhAwAwAAAAkIGAGAAAAEhAwAwAAAAkImAEAQKba7bbm5xc0
   P7+gdrtddnGAiTHSHwAAyEy73daWLVu1uvoaSdL09HZdeeWiNm3aVHLJgGRBjvRnZqeZ2UfN7EYz
   +6yZPa+ssgAAgCON20q8a9clcbC8VVIUOO/adUlu5QSKcGyJ875F0gvc/Vozu6Okz5jZsrt/rsQy
   AQDQeL2txHv3bqWVGI1WWsDs7t+V9N34/5+Z2eck3U0SATMAACU6spVYWl2NnksTMLda27R371at
   rkaPp6e3q9VazLG0QP7KbGE+xMzOkDQj6epySwIAACaxadMmXXnl4qE0jFaLlmlUX+k3/cXpGCuS
   drr7+3pe46Y/YAztdrvrZLWNkxWAkXDjHpoo6aa/UluYzew4SVdI+rveYLljx44dh/6fnZ3V7Oxs
   IWUDqorcQwCTopUYTbCysqKVlZVU7y2thdnMTNKipB+6+wsGvIcWZmBE8/MLWl7erE7uobSoubnd
   2rPnijKLBQBA0ILsVk7SBklPk/RIM9sXT+eXWB4AAADgKGX2krFXjDQIZI471AEAyFbpN/0lISUD
   GA83/QEAMJqklAwCZgAAADReqDnM6DHuMKQAgObi3AHkjxbmQNDnJQBgVJw7gOzQwlwBRw5DGh38
   OjmoADAuWh/rjXMHUIwghsYGAGSPQWwAIBtDW5jN7E/N7ESLvD3uL5mjbcZarW2ant6uaCyXxbgr
   sG1lFwtAhdH6WH+cO4BipEnJeJa7/0TSvKS1kp4u6eJcS9VAnWFI5+Z2a25uN61AAIChOHcAxRh6
   05+Z3eDu55jZ/5K04u7vNbN97j6Te+EadNMfAGSNG8IAIL2J+mE2s8sk3U3SvSWtl3SMpI+6+0My
   Lme/eRMwA8AEGMQGANKZNGCekjQj6cvu/mMzO0XS3d39+uyLetS8CZgBAACQu6SAeWAvGWb2EEmd
   aNUl3dvMJMm6ngcAAABqbWALs5mtKCEwdvdH5lSm7jLQwgwAAIDcTZSSUSYCZmSJXE4AADDIxAGz
   mZ0j6X6Sju885+5/m1kJB8+XgBmZoLcAAHnhxzhQD5Pe9LdD0kZJ95f0fyU9StJed398xuXsN28C
   ZmRifn5By8ubFQ3gIElRv6V79lxRZrEAVBw/xoH6SAqY0wxc8nhJvyvpO+7+TEVdy52UYfkAAKgk
   RlMEmiFNwLzq7rdJutXMTpT0fUmn5VssIFsMH4tJtdttzc8vaH5+Qe12u+zilIZ6ANBEA7uV6/Ip
   MztZ0qWSPi3p55I+mWupgIx1ho89nGfIJVOk13vZfe/erY287E49HK3V2qa9e7dqdTV6HP0YXyy3
   UAAyN1IvGWZ2L0l3KmLQknh+5DADKB058BHqoT9u+gPqYaIcZjPbaGaPMLNHKErFOCn+H0CBuBQO
   AEA50qRkvFiHBzA5XtK5kj4j6bfzKhTKQStJuLgUXi4uu0dCq4cQjlnsm0AzjDxwiZmdJulN7n5B
   PkU6Yl6kZBSErpHCxqXw8oUQnIUglHoI5ZjFvgnUR1JKRpoW5l7fUjSICWrkyK6RpNXV6LmmBgVA
   r02bNrE/KJx64JgFhPMDtgmGBsxm9uauh1OSHqQoJQNAQUK7FA4gwr6JspAOVKw0I/09o+vhrZK+
   5u578yxU17xJyShIKJc3MRgtCcBh7XZbmzc/XQcOvFaStGbNi7V79+Wl5TGzb6JopANlb6KUDHe/
   LPMSITi9/RRv3Phc7dp1iXbtuqSwEwAnnWShXAoHwnGLpLd2/V8O9k2g/ga2MJvZDQmfc3d/YD5F
   OqIMtDCXoIzWZlq4kQd+hNUXrWtoOs6b2Ru3hfmx8d8/if9eLskkPTXDsiFAZdxMww08yFoe+X0E
   4ABCwQi2xRoYMLv71yTJzObd/UFdL11vZvskbc+5bAAwtqx/hHGDTVi42Q4gHahIabqVMzN7WOdG
   PzPboKilGTVVxomIkx9Cx1WQsNC6BqBIaQLmZ0l6h5mdGD/+saRn5lcklK2MExEnP2SNH2H1V+fW
   NdJ/gLCkHumvEzC7+09yLdGR8+SmPwBjyzLo4AYbJGFbA6ov6aa/pF4ynu7ul5tZS1L3m0xRLxmv
   z76oR5WBgBlAMOrc6jds2eq87JPKOsClB5DisX1DGr+XjNvHf++kPgFzRmUDgMqoawrAsBsaueEx
   Gfnt1cb2jTSSesn46/jvjsJKAwAo3LCAj4CwWOTfF4vtG2lMDXuDmf2lmZ1gZseZ2YfNbL+ZPb2I
   wgEAELpWa5ump7dLWpS0GAe428b+vs5N0HNzuzU3tzvz1s52u635+QXNzy+o3W5n9r1AnQ296c/M
   rnP39Wa2RdJjJL1Q0scZ6Q/AIOQDVsuwHFxuQhsu1G2+t1ySWJc92L7RMdZNf10fvtHd729mb5f0
   Hnf/504QnUdhe+ZNwAxUROfEvH//93Tjjf9PBw68VhInn6rgpr/66RcInnXWWdq375nihsIjsX1D
   mjxgvljS70v6paRzJZ0k6QPu/ptZF7TPvAmYa46DVD0ceWJ+q6Q/FidkoFz9ettYu/aVuvnml4v9
   EzjauL1kSJLc/SVm9peSfuLut5nZzyU9LutConm4M7k+jrxpZnfZxQEwwOmn30Wrq9u5oRAY0dCA
   2czuIOm/SbqnpGdLupukX5P0T/kWDXXHncl1tU3S0w494oQMlKNfbxsXXRTti4yqCowmzdDY75D0
   GUm/FT/+tqT3iIAZQKz3xLxmza26//3foXXrTuGEXABSm9BPp7eNfsEx2wgwmjQ5zJ9x94eY2T53
   n4mfy+SmPzP7G0m/J+n77n5On9fJYa6xOt6Z3OTApcnLXqY67kcAUIZJb/r7pKTfkfRJd58xszMl
   vcvdz82gYA+X9DNJf0vA3Ex1CrIIXJAkr22dYZQBIBsT3fQnaYekJUn3MLO/l7RB0jOyKJi7f9zM
   zsjiu1BNdRpqmJxsDMINrmEo+wd62fMHML40vWTsMbNrJJ0XP/U8ScfnWioAqJE8f0wxjHI6Zf9o
   KXv+ACaTGDCb2UMk3VvSTe7+T2Z2mqRXSzpfUa8ZuduxY8eh/2dnZzU7O1vEbIGREbhgEuO2Pibd
   2FU1ebbAln0FqOz5A2UK9erKysqKVlZW0r3Z3ftOknZK+pykd0n6sqRdkr4q6U8lHT/oc6NOks6Q
   dMOA1xyokqWlJZ+bu8Dn5i7wpaWlsouDQCwtLfn09J1dusyly3x6+s5HbB/DXm+CvOtgbu6C+Ls9
   ni7zubkLMvv+0OffBBx/w1Sl41scd/aPVwe+IN3UCYwlrZX0c0lnDHr/uBMBM6qo7gfmui9fGZLq
   dGZmg0vnuXSBS0uNDKbyDijLPmmXPf+6q2r9NuFYW6Ufi0kBc1JKxn+6+y/jqPVmM/uiu38tXbt1
   Omb2LkkbJZ1iZt+U9Ofu/o4s5wFkLe9cxLIvXZFrmY9BN7i2221dd91Nkt4QP7NV3QO/IBtlp66U
   Pf+6q2LKC8faihkUSUv6iaQPdE0/7vp/96DPZTmJFuZUmvALNSR5/loOoZWkSq0BddCvvqemTmnc
   vhzCto/qquJxq4plHkeV9m2N2cL8uJ7Hu7rj7Izi9WCV3cqXFr9Q66WKrSTI3vr1D6jFOh/lOEoL
   LCbBTdfJyoxparNvD4qkQ5hUUgtzlX4NNeUXakcIrel5bh8hrM8qbf91UNf6rutyIVwhnB9GUdQ+
   wr6Ynsa56S+EqayAOYSgJa0qldV9sgNaSDt9XgfmUJaxaieeqgu5vsctW9WOTUAZitj32RfTI2Ae
   UZU2rlACrDR6y7pmzUk+M7Mx9YEi7XoJOfhIo2rlr1p5x9WU5ew2yfGlSsdRoM7YF9MjYB5RlYJQ
   9+qcyI/caZdcWjdSHafZ6XvX3dTUyb5z586+31eVegtZ1faVcTVlOXtNcqJtap0BoWFfTG+igFnS
   r0m6VNKypI/G00eGfS6LqayA2Z1gKg9HnnxHPxGn2enT9jjAASQbTWm5aMpy9pp0uTmOAmFgX0wn
   KWBOHBo79n8kvUXS2yTd1rlXMIP7DYM2qM9UjO/Iu5i/PfLnx73T9uDB+x7V0wS9UQDDTdrzQCjH
   0Sr0elSFMqK6QtkXqyxNwHyLu78l95Kg9roD3v37j9GNN75YBw5Er6U9EQ/b6Vutbfrwh5+sgwc7
   z2xXNAjEVyctPvpICqjqEAB0lmH//h9qzZo/HXl7rbo6dAdVha43q1BGoPEGNT374bSIHZL+m6S7
   Khoie62ktcM+l8UkBi6ptUl7zBj02Z07d/rU1CkeDTXc6ptuQUpGdvqtizrU79E3qZ7qMzMbfGZm
   o8/MbAjy0iaXXY9WdDrNOOugyik/bHOoE02Yw/w1Rc1z3dNXhn0ui4mAGf2kCcbSHMRDOdCHUo4s
   VTkA6Oi3DDMzG4P9IVCHHyl5KHJbHHcdVHV/YZtD3UwUMJc5ETBnL6vgrMwgr6onl37qesKpwzrq
   twxr154Z7HLVoc7zUOQ+Nu46qOpxgG0OdZMUME8NS9kwszVm9nwzu8LM3mNmzzWz4zLMCkFBOnly
   y8ubtby8WVu2bFW73S7te8rQbrc1P7+g+fmFIMp85M2HUQ5jJ1+0ylqtbZqe3i5pUdJinPO7rexi
   jaTfMpx++l3KLlbuQttHJtXJw56b2625ud1B5gZXoYwhq9s2i0ANiqT9cCvv2xWdMX5b0u9IukzS
   24Z9LotJtDBnKqvWgLJbFcZtjQmxFafsusxTHVJNepchxG2oI4uyhbx8oypj+6tT/aURwvKGUAbU
   hybMYb4+zXN5TATM2apLwOxenxtrONhXT8g/BCYtWxn7SB71WeZ+FfL2kYeylzfE4zqqKylgTtOt
   3K1mdh93/5IkmdmZkm7NspUbxZi0T9Wsv2cSdelTsg7ddjVNyNteyGXrJ6/u1MrsZ71q62BSTVte
   NFeagPnFkj5iZl+NH58h6Zm5lQi5ySo4q2qQF0Kg3w8nHISi6H2EAYQwqVCP6xiuan31W9QCPeRN
   ZscrGiLbJX3B3f8z74LF8/U05QPSqtoOmoUmLnOVlb2+ipz//PyClpc3qxMwS9GNb3v2XDHR9/a2
   XE9Pb+dGuhore5/B6ELdR81M7m59XxyUq+GH84j/QNIJ8f8vl/ReSQ8e9rksJpHDjB5l58tlqYhl
   IUe6WvJcXyHuO01bXgCRUHPPNeFNfzfEfx8maUXSYyT927DPZTERMKNbnYK/opYl1INSkaoUOOW1
   vkLed6q0flA+tpd6CPXcNGnAfG3892JJT43/3zfsc1lMBMzoFuoONo6ilqVOdTaOvAPFrE/eea2v
   pm4HBFf1EvIPP4wm1HWZFDCnuenv383sEklzki6O85mHDngCoHxNvyEmz5vK8ujhoez1Vadc0Lx6
   4EB5uEm0PqrYeUCagPkJkh4l6bXu/mMzu6uinjOAQpUdTGSpqGWp4kGpKrI8eXcHqi972XN11VW7
   JWW3vtJsb3ULMAmugLBVrYeoxIDZzI6VdI27n9V5zt2/I+k7eRcM6FWn4K/IZanaQWkcg1pGq/Aj
   6+hANfu7xdNsbwSYCF0V9mfU2KBcDT+cR/x+SacPe18ek8hhRsWFnEMZctlGMSwXLq/lzCoHL21+
   cd7rq255zqHmSGIydTluIUya8Ka/j0v6maSPSPpAPO0e9rksJgJmVFnIJ+yQyzaqMgO9LE7eacpf
   xM2LZ575IJdOcqkVz+cE37lzZ2bzKAPBFYBRJAXMaXKYX555szbQACFf4g65bKHrTf+YdJCNNJeZ
   i7x5UXqBpLMlPVtXXXWNXvayiWdRmiakIwEoxtCA2d1XCigHAIylyLzGPG6MKzs3vzcYj+yWdI6k
   r471nXXqbQNAvipzvBjU9OyH0yIeKulTitIybpF0UNJ/DPtcFpNIyUCFhZz2EHLZxlHUpfey0j8m
   XV9J9dNvmaTzxt4m6rZtoblI6clfaMcLTZjD/BlJ95W0T9Ixkp4p6eJhn8tiImBG1YV8wA25bGmU
   Uf5xAuagTlDBAAAgAElEQVSsyjnu96S5KbL79ampk31mZkOuOdmovqofP4YJLZCrq9COFxMHzPHf
   67ueu3bY57KYCJgB9FPWyWzU+eZZzrQBy8zMBpfOc+kCl5YG3lQ4bvDT+9lRToB1D7rKFnoPMSEL
   LZCrq9DqedKA+WOSbifpckl/KemFkq4b9rksJgJmAL2WlpZ87dozK9EzRl4ng7QBy9LSkk9NnXzo
   fdKdXWplVoaZmY0+NXXKoZ41pqfv7Dt37kxdtroHXWXKs35DC3Ly0IRlDEFox4FJA+YzJE1LOlHS
   Dkmvl3SfYZ/LYiJgLg4tPaiCwwfX8ypxMsvrpJv2e/u9b2rqlIn38d6TXBSIH269TnM8ISDJV571
   W/V1l2b7DC2Qq7OQ4o+kgDlNLxlfM7PbS7qLu+8Y8Z5CVEDdhsTNWmXu4G2Awz063EXdvTpMTb1A
   +/efrXa7HdT6CXFksvXrHzBxHfXvWeMSSZslZdOdG/tduELcrtNKe74ru/eaJqlM94+DImk/3Mq7
   WdIXJH0tfjwjBi6plaq3FuSJVoawHLmtLsUtzWuPSAkIbf3k0XoySkpGHttvFj1rJJWN/W5yRQx2
   E0qr4Cg43yGJJkzJuEbSSZL2dT332WGfy2IiYC4GB5DBqJuw9OvRIQqWm7d+0gYsRQTs4/asMahs
   7HfZqGpQmye2LSRJCpjTjPR3i7v/2My6nzuYRes2wlDly2s4Wl6XskO4RN57mXT//rO1b985qT8/
   yjKEsLxJ0l7GzONy59GXq9811jwqcyl2AmVuR02o31FxvsPYBkXSfriV928kPVXSDYr6Y36zpLcO
   +1wWk2hhLgwtEf1V7dJwXuUNtR4Ol6vl0nk+NXWK79y5c8h7hy9DqMubhxD3/brUf6jLEeI6L1LT
   lx+DacKUjNtLerWkT8fTqyQdP+xzWUwEzMhTmZe081J2rwxl2Llz5xFdpw0KSkZZhjy7gwtpWwo1
   oHMPr67GEeJ+E/I6B8qWFDAPTMkws2lJfyzpPpKul/RQd78lj1ZuIG+9l0Ulpe4ZhMuaYbvqqmt0
   8OAb1OmxYXU16sUhtHUWYm80vb1dhFR37Hf5yHOdh57GBEwiKYd5UdIBSXslPUrS2ZKeX0ShgCz1
   C1TOOuusYAOFSeSVn1eHvL9RlmGU96YNEkIOTusmlMCtDvtNWiH+IAQyNajpWdINXf8fq65eMoqa
   REoGxtR9OTcaGvjIy6JljhSXt6RL2VkOgRyKUXOT0y5D1oMbTHJ5niGO0wttmULbb4rsarAux1Q0
   h8bJYe4NkAmYEYphJ6A0XY/NzGwI6qTaT9Yn2tACiSyVFZSMEiSMW//0pzsaArfh8ljn1DvqIClg
   TkrJeKCZ/bTr8XTXY3f3EyZt3Taz8yW9UdIxkt7m7q+Z9DtRb2ku+/Ve+j54UJqaaungwaj7senp
   7broosVD75XCG8Upj8ubpASUa9yRw7Jab4PSFMgVbp481nmT0k/QUIMi6bwnRUHylySdIek4SddK
   ul/Pe3L6DYGqStOK0e89MzMbK9WKFqWRnOfSBR6NaDd5a01dW4DKbDkvYt5ZrLcqX10YtTU0xGWt
   Wyv+IE1ZTtSXJulWLq9J0kMlLXU9fomkl/S8J686QUWlCR6yPGEWfQJYWlrymZmNLnXSSC5zKepn
   eNLgNsRAIgtl/xDIexvJYr2NmjqSxfJk8T29y75mzUmHfvxWpSvIuu53QB2FGjA/XtKlXY+fJunN
   Pe/Jq05QUWlPPnmcrPM+0fXOLwqUl+I87FMyy2MOJZDISt4Bcwh1NmkZ0tZRVtt8Vt9zZLmXXFpX
   ucCz7B90yFYIx4MiNW15Qw2YFwiYMY6iduCiT3T95helZEQpJegvzx82dWkdTLscWW3z+XxP+qA/
   pBM8AXMktPUyjrocD9Jq2vK6JwfMFr1ePDM7T9IOdz8/fnyhpIPedeOfmZVTOAAAADSOu1u/56eK
   LkiXT0u6r5mdYWZrJD1R0u7eNw2K9JnCnZaWljQ3d4Hm5i7Q0tJS6tdCm5aWljQ9fWdJl0m6TNPT
   d861zL3zm5o6WTMzG4KvpzLXT97b0tzcBfH68Hi6THNzF5S+7FnX4+HtriXphIm3+Sz3nc56npnZ
   oDVrTu37nYfnd16Q66tKx708prrsR3VZDpZ38JSo5II9StIXFPWWcWGf1x3VknQJp8jLOyHduBTy
   /KqqqG2pjEuSRW8DR6cMtHzt2jOD3HcGfefhZWhu+kPZx46k+dclLWVpacnXrDm16ybUU2t9nCYl
   oycmHfRCCFOVA+ayD15lSTowFnXQDGEnb+r6L0qRJ+Ai12UZ224dgpnDy7Dk0c2yzTnBu5d/zBs2
   /7LLl5UoYD7Joy4/z/M1a06q5HKMomnnMgLmgtXl4DCOEALmsgOAo7vCOtVnZjY05oBThFF6fajS
   wb6MbbcOx6sjl6HlU1On+MzMxsotRz9ptuGyj3lpu/us0r7YT9n1jPwlBcxJI/1hTE0eUS1ptKem
   jATVu/4PHJD27XurpM2ZjNiHdNtSHqMlhmzQSH7DjDsCYUiOXoZ3Zr4M49bvpPOsyzbMiJKovEGR
   dAiTKtrC3PRfoUktCUW0MpTdYja4e7jmbQt56mxLMzMb+7bgV3E/HHfbLXubr7uy6rfo/rPHVfb8
   i9KU5WwykZJRLHaq8pV5+a93/UeDLSxVJmirkqR9rV+wkcXNbHkbZ9ud5MdBHS6V562sH1+jzLfs
   9Vj2/IvSlOVsKgLmErBTFS+kOu9u/YxuEqnOj6eQ6nGYpIDi6B8uJ3hnuPEqrIdRjBvQ1fHHfRbb
   b+93lBUw13H9ACEjYEbthXxiGecEnlfQOux7x6nHMgPsYYFMp2xr154ZB8vFBjxFGXf7r2LaSpIs
   jgP9vmPnzp2lHV+q9AMWqDoCZgQn65NAnU78eQX/ab531Hos+4dK2vnXafsYpOhUjhANu+KQpn4G
   fQeBK1B/SQEzvWSgcHW68zuNUe+uz6uXlTy+t+weYdL28NCEHlrG6YWgCfUiZXPMoZcHoNkImFG4
   PIKsUE/8VftxEGo9JkkTyNSh67Q81K1eBm2/oxxzQtwHyujSDkCPQU3PIUwiJaOW8roMHOIl03GW
   tcyUjM770tbjzp07fWrqFI9GvmoFlTuO6shy3+33XaMOrLFz585gjiVlpz0BTSJymJG3UU54od9Y
   luW8Jum9oIyb/kb9ru71ODV1su/cuTOjktZL3ttviD8W0yoiIBw2j5CD0rrlmQMhI2BGZvqdmPMO
   gIs8mWU9r6Tvq3KQ486JPK28t9+Qg700itqOkva3kLflkMsG1A0BMzIx6MSc9wG9yBNGHvPK6kdG
   aA7X1ZJHIxme5zMzG3KbX1V/YNRp/8hDCOUPoQyD1OFYAVRFUsDMTX9IbdCNM0jW76a0snuXyEKr
   tU1XXfUkHThwrKTXSZJuvPHFarfbmS9H1W6eRHoh3GQXQhkGqduNmUBlDYqkQ5hEC/PI8myFS+qf
   tC6XnIuaV8gtWqOYmdlYyHJUub7yTstJk58best8CGUMoQxNQ50jNCIloxnKDFzrdFNTEfMq+kdA
   0T+islblgNl9vLScUfP8+72Xy/kIFdsmQkTA3BBF5JQWGRiH1LVTHuoQmBd10qvjyXXYqHRZLG/V
   f2iMIo/9iRbQ/DRp20R1EDA3RHQAarl0ZFdfVTnQ9wYJ0gnx8tQjQCpDESelrIKKYd9Tt+Alad1k
   td6aEpTk8YOqjj/SQtKUbRPVQsDcEEtLS/EgEtU8CPU7gEYt5dVajpBU5aTUxOAkaZmzWm9Vq9dx
   fxTlsZ1XZd+pqqptm2iGpIB5qoT7DJGTTZs2af36B5RdjEZrt9uan1/Q/PyC2u122cVRq7VN09Pb
   JS1KWozv/t9WdrGOcmSvIVFvGHn2wBLCeur0fjA3t1tzc7uP6PUjq/WWNI+0iqqrTk8oy8ubtby8
   WVu2bA1iH0I+stg2gUINiqRDmEQL88iq/Ku96ikZZdd90o1foacyFNmaV/Z6SiuE9VlkXU2yDZCS
   ASALIiWjfpJOmlUIkAap8k1/ZV7CrfrJvSqBWdmKXs9VGjQojx8YVT6WAhgdAXPNFN0l2czMRl+7
   9kyfmdnASSNBmYFYlYPAjqKCkyrXVdqyZ1WXVW/5r/oPSQDFImCumaJOYktLS75mzaldKRLrfM2a
   kzjhDFDmybnKQeAkxgkMi+hqr8x+r7NcvqK36azrblh9NbEFuYnLDKRFwFwzZQ4WIZ3XiEBsXGWd
   jJrYkjbJMue1nkLo9zrr40OVA6wi+rqukiYuMzAKAuaaKeqgR8BcLVUObMYRYqt6HmXqXa/D1nOI
   9VKWIrruS5p3aPsj2waQLClgPrboXjkwuU53PJ1ut1qtfLrjabW26aqrnq4DBzrPvEhr1tyqVmtH
   5vPC5DZt2lS7bpna7XbXdr6tdss3TKertajLPWnv3q268spF7dlzxcDPtFrbtHfvVq2uRo+jLukW
   iyhucIo6VvYatN6atv0CtTIokg5hEi3MpeOmP5Rl2JWUEC8vZ12mcVsEQ+pFJ8SWVvd8t59QW3JD
   3GeAkIiUDABVk/YGt9CCsSzLlEc+ctE38YUcoOW1/YQaMLuHuc8AoSBgBgfJBipznWcx75CDjqKE
   0mKdtqy967yp6zD0HwoA+ksKmMlhbgDy6ZqnzHWe1bzJxS0vB3dUg9Z5U1VlvQEYwaBIOoRJtDBn
   oqmtPE1WZtdiWc67ildGQi5zXi2fg9Z5ES2tIdc3gGoRLcwAxtVut7V585N04MBZkqSrrnqSdu9+
   dyEtZlXr+SP0qzlFt3zmPb/Q6xtAjQyKpEOYRAtzJpJaeerWOlPU8oReb1m27M3MbHBp3aHvktb5
   zMyGQuZdNf1aWmdmNga9rWShrHXO1TNkLfRjO/IlbvpDv4NA3QKbopanKvU27MCf9sSwdu2ZRwUl
   a9eeOdG88/ps2foFcFNTpwS/rWShjPVWxYC5ytt33VXl2I78EDCjryqebJKUOWR41eptlBPDzMzG
   vq2mZZcrRL3ln5o62aVWpbeVkPXW95o1p/rMzIZgg9Gqb991V4djOyaTFDBPlZYLAqA0u3ZdEud9
   bpUU5YB28kx7XXTRhVqz5sWSFiUtas2aF+uiiy486n3tdlvz8wuan19Qu93OvVwh6uTszs3t1tzc
   bq1ff7akc8ouVuGy2BbS6K7vmZl3SLpF+/Y9W8vLm7Vly9Zc5z2Oqm/fQKMNiqRDmEQLc67q1tpB
   SkZ6o7akpEnvyKJO6tbCE9q2UkQ6APnMg1WhjE0W2v6K4omUDAxSt3w6bvpLJ9QBMUYtVxXWQyhl
   LCoYKCsozHpkyDzWGwFZ+ELZX1EOAmaggbK66S+NMvpeJvgYTZ1z/JeWlnxmZmN8g2Wr7/YwyvaS
   57ZFQAaEi4AZaJiig8mdO3fGN7gVF7xyeXs0RdVX0dtevxstZ2Y2HDXPUZa/LtsWwTkwmqSAmYFL
   gBo68uYiaXU1ei6PAR3a7bZe9ao36+DBZ0l6q6amvqiXvewFjRw8ot1udw3SsW2sOsjiO/opaqjx
   ogdH6d3WDx6U1q3b3cjtrxuDugAZGxRJhzCJFmZgLEW2kJXVGhdaSkYW5cl7merY4ph2+wslJaMo
   dWklB4qk0FqYzewJknZIOkvSb7j7NWWUA6iroloTy1R0S+YwWbTq531loGpDjaeRdlsfZXsJbdsC
   UL6yUjJukLRF0l+XNH+g1oo84ZcZnNcxAEQkbWrKqIFw2u2l6ttWE340A0WyqAW6pJmbfVRSa1AL
   s5l5meUDkE5eebdV0pszOj29feSc0Sy+ow6oh2ywXwKjMTO5u/V9jYAZQJWFFBSEfNPfOMoqy/z8
   gpaXN6uTmiJFo/nt2XNFIfMH0ExJAXNuKRlmtizpLn1eeqm7fyDt9+zYsePQ/7Ozs5qdnZ24bADq
   IbSeALK4jB9KKkBodQsAWVtZWdHKykqq99LCDNRMSC2UeaMlMnud7eczn7lON9/8+5JeF79SXN2S
   kgGgDEktzFNFF6aPvgUDMLpOoLG8vFnLy5u1ZctWtdvtsouFgrXbbc3PL2h+fmGk9d+9/dx888sl
   LUoqfvvp3Mg3N7dbc3O7CZYBlG9Qf3N5Top6yPimpFVJ35X0zwPel1XXekhQx75Zy1J2XTat79U6
   9JebtUnqpN/2I51H3SIzZR8jgSRK6Ie5lBZmd7/S3U9z92l3v4u7P6qMcoAWySxRl8WjJfJoR/bl
   HKU1dFJ0xrF27Q+o24YZ9wrFsM9yjESlDYqkQ5hEC3Nq4/5qb1qLZJ5CqEtaXKslj9a2SbbD3u1n
   aupkn5nZwDbUIJMcQ4Z9NoRjJJBEobUwI1v8akcHLa7Vkdd+22pt0/T0dkX5x4vxgBXbUn22s/3M
   zLxDU1MtHTz4LO3b92yOKQ0yyRWKrK9uACEpa6Q/ZGiS4XQZDSo7odRlKN2SIVlew2BPOsrjpk2b
   tGvXJTp4cFfmZZOa1YsLjhTKMRIYBwFzwxU5hHLdbdq0SS972XP1+te/UpL0whc+l7pEKUL90UTf
   zuGbJKgd9lnON6i0QbkaIUwihzkV8lbDwHrAKELeXvIqGzms1TBJbj29YGCQKmwbSshhLnXgkmEY
   uCQ9LnOWj0E0MKqQ99tRy5bm/ewjQDNVZTCipIFLCJiBjBAMoKnSngyrctIEkK2qnB9DH+kPqIVJ
   eifI2iT9qAIdabejtL0j0IsLgKripj8gI6Hc0MKNVeMJOT0iT4OWO6/tKNQbEgHkpw49pJCSAdRM
   VS59haSpqQJJyz3KdtTU+gOQXhUaJZJSMmhhBtB4efWJHLqsljuUqyuTqsIJHaiqql9dImAGaqYO
   l75QvlG3o6qfDItOZSI4B6qFlAyghkY5GXPibm5KwbDlbtK2UWQqU1O3NyB0dCsHoC9O3Ic1KTjs
   1tTl7lVkwMx9BkCYyGEG0FdTc3f7CSmloMggNqTlLhOpTACSEDADQEDoFrAcRd64SHAOVA8pGUCD
   kZIRHi7XNwOpMEB4SMkAKqiIE2pdugMDqoZUGKBaaGEGAkTLb3Ox7gGgHPSSAVQMl+Wbjcv1AFA8
   UjIAoEK4XF+8cX6k8MMGaI6psgsA4Git1jZNT2+XtChpMb6LflvZxUKG2u225ucXND+/oHa7XXZx
   +qpCGbPQSYNZXt6s5eXN2rJl69DlHeczAKqLlAwgUKG0XoVSjjqpQp5yFcqYlXFSoEibAuqHlAyg
   gkK4LF/lPoFDDvSrMGBMFcoIAEUhYAYwUFWDpioH+ijeOAOJ9PvMxo3P1fz8wqHX2d6A+iBgBlA7
   F174Sq2u3kvSbknbtLr6mqAC/SqM9FaFMmZlnP7Iez+zceNz9apXvZkfaUBNkcMM4CiddIb9+3+o
   G2+8TgcOvFFSNfJY2+22Hv3oJ+vgwTfEz2yX9DTNzX01qPzSkFNGOqpQxlCQ0wxUHznMAFLrTWdY
   s+bFmpm5VOvW3bkSIwHu2nVJHCxvPfTc1FRLrdY7yytUHyHkqA9ThTICQBEImAEcoTdv+cABad26
   areUrV//AAI/5KpJKSxAExEwA6iVfoHLRRcRuCBf4+RBA6gOcpgBHNJut3Xhha/UddfddCgHuAp5
   y73IvQUAjCoph5mAGQhYkYHfkbnLN2hq6jKtX/8AXXTRhQScQMXxIxIYjoAZqKCiR1rjLn+gnpo0
   aiMwiaSAearowgBI58ib77Ye6ksYCEW73db8/ILm5xfUbrfLLg4G4FgCTI6b/gBI4i5/jIbRFAE0
   CQEzEKiiA1ju8scoqjpsehPxYxiYHAEzEKgyAlgGqgDqhx/DwOS46Q9AY9BTQHa4kQxA3dBLBoDG
   I8DLHj9AANQJATOAxqPbPABAErqVAwAAAMbETX8AGoGeAgAA4yolJcPMXivpMZIOSPqypGe6+0/6
   vI+UDACZIecWADBIcDnMZjYn6cPuftDMLpYkd39Jn/cRMAMAACB3weUwu/uyux+MH14t6R5llAMA
   ssRQ0QBQTyHkMD9L0rvKLgQATIKhogGgvnJLyTCzZUl36fPSS939A/F7Xibpwe6+MOA7SMkAUAl0
   WwcA1ZaUkpFbC7O7zyW9bmbPkPRoSb+T9L4dO3Yc+n92dlazs7OTFw4AAACNtrKyopWVlVTvLeum
   v/Ml7ZK00d33J7yPFmYAlcBIggBQbSH2kvFFSWsk3Rw/9S/u/id93kfADKAy6LYOAKoruIA5LQJm
   AAAAFCG4buUAAACAqiBgBgAAABIQMAMAkBKD0wDNRA4zAAAp0BMKUG/c9AcAwIQYnAaoN276AwAA
   AMaU20h/AADUSau1TXv3btXqavR4enq7Wq3FcgsFoBCkZAAAkBKD0wD1RQ4zAAAAkIAcZgAAAGBM
   BMwAAABAAgJmAAAAIAEBMwAAyAQjIaKuCJgBICWCAWCwzkiIy8ubtby8WVu2bGU/QW3QSwYApMCw
   yEAyRkJE1SX1ksHAJQCQwq5dl8TBchQMrK5GzxEwA0D9kZIBpMCleABI1mpt0/T0dkmLkhbjkRC3
   lV0sIBOkZABDcCkeEtsBkAYjIaLKGOkPmAB5eeggGACA+iKHGQAysGnTJoJkAGggAmZgiFZrm/bu
   3arV1ehxlJe3WG6hAABAYUjJAFLgUjwAAPVGDjMAAACQIClgpls5AAAAIAEBMwAAAJCAgBkAAABI
   QMAMAAAAJCBgBgAAABIQMAMAAAAJCJgBAACABATMAAAAQAICZgAAACABATMAAACQgIA5pZWVlbKL
   UEnU2/iou/FRd+Oh3sZH3Y2HehsfdTeeceuNgDklNszxUG/jo+7GR92Nh3obH3U3HuptfNTdeAiY
   AQAAgBwQMAMAAAAJzN3LLsNAZhZu4QAAAFAr7m79ng86YAYAAADKRkoGAAAAkICAGQAAAEhAwNzD
   zM43s8+b2RfNbHuf188ys38xs1+aWauMMoYoRb091cyuM7PrzewTZvbAMsoZohR197i47vaZ2WfM
   7LfLKGdohtVb1/t+w8xuNbMLiixfyFJsc7Nm9pN4m9tnZv+jjHKGJs02F9fdPjP7rJmtFFzEYKXY
   5l7Utb3dEO+zJ5VR1pCkqLd1ZrZkZtfG29wzSihmkFLU3clmdmV8fr3azO6f+IXuzhRPko6R9CVJ
   Z0g6TtK1ku7X855TJf26pJ2SWmWXOYQpZb09VNKJ8f/nS/rXsssdwpSy7u7Q9f85kr5UdrnLntLU
   W9f7PiLpnyQtlF3uEKaU29yspN1llzWkKWW9nSTpRkn3iB+vK7vcIUxp99eu9z9G0ofKLnfZU8pt
   boeki+L/10n6oaRjyy572VPKunutpJfH///asG2OFuYjnasoGPmau98i6d2SHtf9Bnf/gbt/WtIt
   ZRQwUGnq7V/c/Sfxw6sl3aPgMoYqTd39vOvhHSXtL7B8oRpab7HnSnqPpB8UWbjApa27vneKN1ia
   enuKpCvc/VuS5O7sq5G021zHUyS9q5CShS1NvX1H0gnx/ydI+qG731pgGUOVpu7uJ+mjkuTuX5B0
   hpmdOugLCZiPdHdJ3+x6/K34OSQbtd7+SNIHcy1RdaSqOzP7fTP7nKR/lvS8gsoWsqH1ZmZ3V3SA
   fEv8FF0CRdJscy7pt+JLlR80s7MLK1240tTbfSWtNbOPmtmnzezphZUubKnPEWZ2e0mbJF1RQLlC
   l6beLpV0fzP7tqTrJD2/oLKFLk3dXSfpAkkys3Mlna6ExrxjMy5g1XFCHU/qejOzR0p6lqQN+RWn
   UlLVnbu/T9L7zOzhki5XdPmoydLU2xslvcTd3cxMtJh2pKm7aySd5u6/MLNHSXqfpF/Nt1jBS1Nv
   x0l6sKTfkXR7Sf9iZv/q7l/MtWThG+Xc+lhJe939x3kVpkLS1NtLJV3r7rNmdqakZTNb7+4/zbls
   oUtTdxdLepOZ7ZN0g6R9km4b9GYC5iP9u6TTuh6fpuhXCZKlqrf4Rr9LJZ3v7j8qqGyhG2mbc/eP
   m9mxZnaKu/8w99KFK029PUTSu6NYWeskPcrMbnH33cUUMVhD6677ZOvu/2xm/9vM1rr7zQWVMURp
   trlvStrv7quSVs3sY5LWS2p6wDzKce5JIh2jI029/ZakV0mSu3/ZzL6qqEHl04WUMFxpj3PP6jyO
   6+4rg76QlIwjfVrSfc3sDDNbI+mJkgadXGmtOmxovZnZPSW9V9LT3P1LJZQxVGnq7sy4hVRm9mBJ
   aniwLKWoN3e/t7vfy93vpSiP+TkEy5LSbXN37trmzlU0yFWTg2Up3fnh/ZIeZmbHxKkFvynppoLL
   GaJU51YzO1HSIxTVI9LV2+cl/a4U7beKguWBQV+DpDnOnRi/JjN7tqSr3P1ng76QFuYu7n6rmf13
   SW1Fd1i+3d0/Z2b/NX79r83sLpI+pSi5/qCZPV/S2UmVXHdp6k3Sn0s6WdJb4vPwLe5+blllDkXK
   uluQ9IdmdouknylqgWm0lPWGPlLW3eMlPcfMbpX0C7HNpao3d/+8mS1Jul7SQUmXunvjA+YR9tff
   l9SOW+gbL2W9vVrSO8zsOkWNoH/Gj9vUdXe2pMvMzCV9VtH9VQMxNDYAAACQgJQMAAAAIAEBMwAA
   AJCAgBkAAABIQMAMAAAAJCBgBgAAABIQMAMAAAAJCJgBIAdmdpuZ7Yuna8zsdDP7RPza6Wb25K73
   ro+HoB51Hitm9pCe515hZq/uee5BZjawP2Az22FmrVHnDwBNQcAMAPn4hbvPxNOD3f3r7r4hfu1e
   kp7S9d4ZSY8eYx4eT93+XtGoVt2eFD+f9D0AgAEImAGgIGbWGRH0YkkPj1uf/0zSX0h6Yvz4CWZ2
   BzP7GzO7Om6d3hx/ftrM3m1mN5nZeyVNS7Luebj7FyX9KB7SuuMJkt5lZs82s38zs2vN7D1mNt39
   0Xgeh1qtzWydmX01/v8YM3tt/PnrzGxb/PxdzexjcdlvMLOHZV1vAFA2hsYGgHxMm9m++P+vuPuC
   Dolfx3YAAAI8SURBVLfkbpf0Ind/rCSZ2fckPcTdnxc/frWkD7v7s8zsJElXm9mHJP2xpJ+5+9lm
   do6ka9S/dfhdilqV/83MzpN0s7t/2cx+5O6XxvN4paKhYP+q57P9Wq0Vv/fH7n6umd1O0l4z2yPp
   AklL7v5qi8a9v8OI9QQAwSNgBoB8rLr7zIDXrM/j7ufmJT3WzF4UP76dpHtKerikN0mSu99gZtcP
   +P5/kPTJOC+5Ox3jHDPbKelESXeUtDTC8szHn398/PgESfeR9ClJf2Nmx0l6n7tfN8J3AkAlEDAD
   QPn6teheEKdXHBI14B4VbB/9Ze7filMpZhW1AJ8Xv3SZpM1xsL01fr3XrTqcrnd8z2v/3d2Xez9g
   Zg+X9BhJl5nZ69398mFlBIAqIYcZAIr3U0l3SnjclvS8zgMz67RUf0zxzYJm9gBJD0yYx7skvUHS
   l9392/Fzd5T03bg1+Gk6HKh3t3B/TdKvx/93WpM7ZfoTMzs2nv+vmtntzeyekn7g7m+T9DZFNzAC
   QK0QMANAPvq1Gneeu07SbfHNd8+X9FFJZ3du+pP0SknHmdn1ZvZZRTcFStJbJN0x7iLuLyR9OmH+
   75F0tqLAuePlkq6WtFfS53rK1Snb6yQ9x8yukXRK1/Nvk3STpGvM7Ia4LMcqaqW+Nn7/HyhOGQGA
   OjF3ehMCAAAABqGFGQAAAEhAwAwAAAAkIGAGAAAAEhAwAwAAAAkImAEAAIAEBMwAAABAAgJmAAAA
   IAEBMwAAAJDg/wM+aPuY3JAg4wAAAABJRU5ErkJggg==
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h4 id="Histogram-of-standardized-deviance-residuals-with-Kernel-Density-Estimate-overlayed">Histogram of standardized deviance residuals with Kernel Density Estimate overlayed<a class="anchor-link" href="#Histogram-of-standardized-deviance-residuals-with-Kernel-Density-Estimate-overlayed">&#182;</a></h4>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <p>The definition of the deviance residuals depends on the family. For the Binomial distribution this is </p>
   <p>$$r_{dev} = sign\left(Y-\mu\right)*\sqrt{2n(Y\log\frac{Y}{\mu}+(1-Y)\log\frac{(1-Y)}{(1-\mu)}}$$</p>
   <p>They can be used to detect ill-fitting covariates</p>
   </div>
   </div>
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[41]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">resid</span> <span class="o">=</span> <span class="n">glm_mod</span><span class="o">.</span><span class="n">resid_deviance</span>
   <span class="n">resid_std</span> <span class="o">=</span> <span class="n">stats</span><span class="o">.</span><span class="n">zscore</span><span class="p">(</span><span class="n">resid</span><span class="p">)</span> 
   <span class="n">kde_resid</span> <span class="o">=</span> <span class="n">sm</span><span class="o">.</span><span class="n">nonparametric</span><span class="o">.</span><span class="n">KDEUnivariate</span><span class="p">(</span><span class="n">resid_std</span><span class="p">)</span>
   <span class="n">kde_resid</span><span class="o">.</span><span class="n">fit</span><span class="p">()</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[42]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">,</span> <span class="n">title</span><span class="o">=</span><span class="s">&quot;Standardized Deviance Residuals&quot;</span><span class="p">)</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">hist</span><span class="p">(</span><span class="n">resid_std</span><span class="p">,</span> <span class="n">bins</span><span class="o">=</span><span class="mi">25</span><span class="p">,</span> <span class="n">normed</span><span class="o">=</span><span class="bp">True</span><span class="p">);</span>
   <span class="n">ax</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">kde_resid</span><span class="o">.</span><span class="n">support</span><span class="p">,</span> <span class="n">kde_resid</span><span class="o">.</span><span class="n">density</span><span class="p">,</span> <span class="s">&#39;r&#39;</span><span class="p">);</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAsEAAAGtCAYAAAD3b6nIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xd81dX9x/HXYToQRKRaEUUUFRW1tuIuuEGqotYBKO7R
   un6OtlrbirV2W6t178VyFfdWrBu3qIigoiIqQ0UUZeX8/vjeYAghuUm+yfeO1/PxuI8k9968v5/c
   JDefnHvO+YYYI5IkSVI5aZF1AZIkSVJzswmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJ
   UtmxCZZUUEIIfUMIH6WYd1gI4ckqH88JIXRLKz+XOTaEcGSamfU49n0hhEOyOHZzqu37Vv173Mjj
   TAkh7JxGlqTCZhMsaSkhhO1DCM+EEL4MIcwKITwVQvhJ7rbUGo4sxBhXijFOSTs2d1lKCGFYCGFB
   COGr3GViCOE/IYTVUzlwjHvEGG9KIysNVb7eObmfn+dCCDs0NreJvm81HoplfC8llRabYElLCCG0
   B+4BLgQ6Al2Ac4B5WdaVjxBCy6xrqEEERsYY25M8nvsAqwMvpdUIF5jKr3cloBPwCHBbtiVJ0tJs
   giVVtz4QY4yjY+K7GOPDMcbxIYSewGXANrmRvs8BQggDQgivhBBmhxA+DCGcXRkWQugWQqgIIQwN
   IXwQQpgRQvhtlduXDyFcH0L4PITwJrBl1WJCCGeEECbnRlHfDCEMrHLbYSGEp0MI/wohzATODiGs
   EkK4K1fL88C61fIqQgjdQwhr5L6GysvcEEJFlfsdEUJ4K1fXAyGEtarctmsI4e3cSOd/gJC71GTx
   bTHGRTHGt4ADgRnAaVUyfxZCeDWE8EXua+qVu/43IYRbq30NF4YQLsy9v3gqRghh3RDCYyGEmbnH
   +eYQQocqnzclhHBaCOG1XO2jQghtq9y+d66G2bnHfPfc9R1CCNeEEKaFEKaGEM4NISzr78cSXy8w
   AugcQli1rqwQwnohhCdytc0IIYyq/n3Lvd9pWd/jKj9vLapcl/djVO1x7h1CeDF3nE9DCOcv42uW
   VIRsgiVVNxFYlGtM+4UQOlbeEGOcABwHPJt7eXqV3E1fAwfHGDsAA4BfhBD2rpa7HUmDvTPwhxDC
   BrnrzwbWAboDuwOHsuTL0ZOB7XMjqecAN4cQVqtye2/gXeAHwJ+BS4G5JKOtRwCHU8PL2zHGabmv
   YaXcqOUdwEhImkHgTJJR21WBJ6vctipwO/BbkpHOd3NfW94voccYK4A7gR1ymT8CrgGOBlYBrgDu
   CiG0BkYBe4QQ2uXu2xLYHxheGVft2OcBPwR6Al2BYVUPnfvc3Uke802Bw3K5vYEbgNNy38efAlNy
   n3c9MJ+k2fwRsBtwVF1fZwihDTAUeDfGODOPrHOBB2KMK5O8AnHRMqIvIY/vcbWvO9/HqKoLgQty
   j0d34JZajiGpyNgES1pCjHEOsD1J03AVMD2EcGcI4Qe5uyw14hljfCLG+Gbu/fEkjVufanc7J8Y4
   L8b4OvAasFnu+v2B82KMX8YYp5I0HouPEWO8Lcb4ae79W4BJwFZVcqfFGC/JNZYLgH2BP8QYv83V
   dENNNVcVQvgNsAFJQwVJo/+XGOPEXO5fgM1zo8F7AG/EGO/Ijez+G/i0tvxl+ISk4QU4BrgixvhC
   bvT9RpLpJ1vHGD8AXiZpyAF2AubGGMdVD4wxvhtjfDTGuCDXdF7A0t+Hi2KMn8YYvwDuBjbPXX8k
   cE2M8dFc1rQY48TcPxz9gVNyj+kM4N/AQbV8bQeEEL4gaVSPInnMyCNrPtAthNAlxjg/xvhM9eDc
   PwH1/h7X8zGqNB/oEUJYNcY4N8b4fD7HkFQcbIIlLSXG+HaM8fAYY1dgE2ANkmalRiGErUIIj4cQ
   pocQvgSOJRklrapqozgXaJd7fw2g6m4QH1bLHhqSqRZf5BqrTaplV/3czkCr2vJqqL0/cBIwMMZY
   Oe95beDCKseclbu+C8kI4tRqMQ3ZzaJLldy1gdMqj5c75pokjw0kUwoG5d4fzPejwNW/ltVyUxym
   hhBmAzdR+/fhW2DF3PtrkoxqV7c20Br4pEptl5M81ssyOsbYEVgNeAM4Mc+sX5M0s+NCCG+EEA6v
   Ibve3+Oq8nyMKh1J8urFhBDCuBDCgHyPI6nw2QRLqlWMcSLJSNsmlVfVcLcRwBhgzdxL2ZeT//PL
   J8BaVT6uOvd2beBK4HhglVxj9QZLjvpVrWcGsHBZedXlpmRcD+wfY/y4yk0fAsfEGDtWuawYY3w2
   V2/XKhmh6sc1WOrxys1X3ZNkmkXl8c6rdrx2McbRudtvA/qGELoAA0ke75r8GVgEbJJ7Cf8Q8v8+
   fASst4zr5wGdqtTWIcbYaxk5ke/nBM8iGeU+JoSwTl1ZMcbPYozHxBi7kPwjdWnlPOAq6voef5N7
   u0KV66ouQMz7MYoxTo4xDo4xdgb+BtwWQlh+GV+3pCJjEyxpCSGEDUIIp+YaLkIIXUlGIZ/N3eUz
   YM3cfNVK7YAvYozzc3NLB5P/HNlbgDNDCCuHENbk+1FDSEYpIzATaJEbGdykhgxg8UKsO4BhIVlw
   txHJHOOavs72JPNyz6rhZffLgd/mPr9yMdf+udvuAzYOIewTQmhFMopc2y4Pixv2EEKrkCwuHEky
   h/lfuZuuAo7LLcQKIYQVQ7LYsF3u65oBjCVp2N/L/WNSk3YkTeBXue/fr2qpq3p91wCHhxB2CiG0
   CCF0CSFsEGP8BHgI+FcIYaXcbeuGEH5a19ebq/0dkmkXv64rK4Swf+5nAOBLku99RbW8Wr/Hucfq
   Y+CQEELLEMIRLLk4Mu/HKIRwcAihcpR6dk31SCpeNsGSqptDMuf2+RDC1yTN7+t8v5PBo8CbwKch
   hOm5634J/DGE8BXwe2D0kpG1NsTnAB8A7wMPADdW3j+3k8L5uRo+JWmAn6qWWz37BJJG51Pg2twl
   VvscgC1IXuq+IHy/Q8RXueOOIRn5G5V7yXw8yWIycvNI9wf+StKcr1etpuoicGAIYQ5JY3cnyWjm
   j6vMdX6JZFHcxcDnJPOeh1bLGUGyqHBZo8CQPJZbkDRsd5Ms4MtrwViM8QWSBWYX5Oocy/cjrEOB
   NsBbufpuZdmNf03fk38AQ3PzymvL+gnwXO6xuhM4KX6/N3DVzLq+x0eTNLczgY2Ap6vcVp/HaHfg
   jVw9FwAHVZkyI6nIhRhrH6wJIfQjmQvYErg6xvi3arf3JXmyei931e0xxj+lX6okSZKUjla13Zhb
   hXsxsAvJy0svhBDuym2TVNUTMca9mqhGSZIkKVV1TYfoDUyOMU6JMS4g2fao+t6fkOfWNJIkSVIh
   qKsJ7sKS29BMzV1XVQS2DckZiO6rXEgiSZIkFapap0OQ3+rul4GuMca5uf02x5AsNpEkSZIKUl1N
   8Mcsuf9lV6ptEp87u1Tl+/eHEC4NIawSY/y86v1CCHmfUlSSJElqjBhjrdN162qCXyQ5ZWQ3YBpw
   IN+ftQhYfBrM6THGmNsfNFRvgKsUk2fZKhfDhg1j2LBhWZehAuPPhWriz4Vq4s+FapKcx6h2tTbB
   McaFIYQTgAdJtki7JsY4IYRwbO72K4CfA78IISwkORVqbeeTl6TFanuSOuecc+qV5T/ZkqT6qGsk
   mBjj/cD91a67osr7lwCXpF+apPJQU/M6LHfJlxvUSJLqxzPGKVN9+/bNugQVpL5ZF6AC5POFauLP
   hRqqzjPGpXagEKIvV0qqKpkOkcbzQnA6hCRpsRBCnQvjHAmWJElS2bEJliRJUtmxCZYkSVLZsQmW
   JElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYk
   SVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJ
   UtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS
   2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZ
   sQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmx
   CZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJ
   liRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZsQmW
   JElS2bEJliRJUtmxCZYkSVLZsQmWJElS2bEJliRJUtmxCZYkSVLZqbMJDiH0CyG8HUKYFEL4TS33
   2zKEsDCEsG+6JUqSJEnpalXbjSGElsDFwC7Ax8ALIYS7YowTarjf34AHgNBEtUoqdu+8AyNHwpNP
   wrvvMhP4hrV4j+48x9aMYSDPsxU+jUiSmlpdI8G9gckxxikxxgXAKGDvGu53InAbMCPl+iSVgscf
   h113hR12gK++glNOgUceYUPgp/yP8ziLBbTmRobyMluwKw9lXbEkqcTVOhIMdAE+qvLxVGCrqncI
   IXQhaYx3ArYEYpoFSipi06fDaaclI7/nngsHHABt2y6+eSYwk258QDceYVf+wB8ZyBgu4Xgm0YNj
   uYKpdM2ufklSyaprJDifhvbfwBkxxkjyGqavY0pKGt/NNoPVV4c334RDDlmiAa5ZYAz7sDFv8hTb
   M47ejgpLkppEXSPBH8MSwzBdSUaDq/oxMCqEALAq0D+EsCDGeFf1sGHDhi1+v2/fvvTt27f+FUvK
   VO53vVZHAn8GDgEe+uc/4Z//rNcxFtCGv/BbnmY7RjCY8zmNCzi1QfVKkkrf2LFjGTt2bL0+JyQD
   uMu4MYRWwERgZ2AaMA4YVH1hXJX7XwfcHWO8o4bbYm3HklQckiZ42b/LZ/AXjuBaBnAvk1i/rrRa
   swC6MJVH2ZnhDOFcfk/NLzYFfH6RJFUKIRBjrHXUptaR4BjjwhDCCcCDQEvgmhjjhBDCsbnbr0it
   WklF71f8ncO5jj48wSeskUrmx6xJH57gEXahDfP5PX9KJVeSVN5qHQlO9UCOBEslYVkjwcdzMSdx
   ETvyONPokm9ajVk16cRMnmY7LuRkLuOXS+X4/CJJqtTokWBJykd/7uMszmNbnqlHA1w/s1iV/tzP
   U2zPR3TlHvZskuNIksqDI8GS6qX6SPAmjOdRdmYgY3iWbeubRn13VdyScdzLAHbiMd6g1+Icn18k
   SZXyGQmu87TJkrQs7ZnNGAZyChc0oAFumBfozWmcz238nHbMaZZjSpJKjyPBkurl+5HgyCgOYhad
   OJ5LG5pGQ8+vcwXH0IHZHMQooIUjwZKkxRwJltRkjuFKNuRtTuVfmRz/JC5ifd7hGK7M5PiSpOLm
   SLCkegkhsCFv8T9+yvY8xTts0Jg0GnOm9Q2ZwJPswNbMYrLPL5KkHEeCJaWuBXAdh3M25zSyAW68
   t+nJn/gd1wMsWpRpLZKk4mITLKleTgXmsgKXc1zWpQBwESexAOD887MuRZJURJwOISl/b7/NzJ49
   2ZL3mMI6KQQ2bjpEpW4E3u/UCcaNg+7dG1+WJKmoOR1CUnpihF/8gj9CSg1weqYA/OpXcMIJSZ2S
   JNXBJlhSfoYPh9mzG7wZWpM79VT48EO47basK5EkFQGnQ0iq25dfQs+eMGYMYeutSWMKQyKd6RCL
   zxj31FNw0EHw1lvQvn0KuZKkYpTPdAibYEl1O/FEWLAALr98qdMmN07KTTDAUUfBCivARRelkCtJ
   KkY2wZIa7+23YYcdkredOhV+EzxrFmy8MTzwAGy+eQrZkqRi48I4SY3361/Db34DnTplXUl+OnWC
   YcPglFNcJCdJWiZHgqUykYzg1k9f4BqgJzB/iVsKeCQYYOHCZBT4T3+CgQNTyJckFRNHgiVVE/O+
   BBZxPj/iDEYzf4nbikCrVvCvfyXbps2bl3U1kqQCZBMsqUYHczPzaMut7J91KQ2z226w/vpw8cVZ
   VyJJKkBOh5DKRH0WtC3Ht7zD+hzIaJ5l2+pJeefkUVVKWdWmQ1SqXNT31lvQuXMKx5EkFQOnQ0hq
   kF9wGS/ykxoa4CKz4YYwaBCcc07WlUiSCowjwVKZyHckeAW+4V3WZTceYjyb1pSUV06eVaWUtYyR
   YIAZM5Jm+IUXoHv3FI4lSSp0jgRLqrcTuJgn6LOMBrgIde4MJ50EZ5+ddSWSpALiSLBUJvIZCV6J
   r5jMevRlLBPYaFlJdebUo6qUsmoZCQaYMwfWWw8eeQR69UrheJKkQuZIsKR6OZH/8BC71dIAF6mV
   VoIzz4Szzsq6EklSgXAkWCoTdY0Ed+BLJtGD7XiaSaxfW1KtOfWsKqWsOkaCAb77DjbYAEaMgO22
   S+GYkqRC5UiwpLydzIXcy4A6GuAittxyyemUzzzT0ylLkhwJlspFbSPB7ZjDe3TPYxQYinYkGJLT
   KW+6KZx/PvTvn8JxJUmFyJFgSXk5hit5nB1LdxS4UqtW8Kc/JaPBFRVZVyNJypBNsFTm2jCPU/kX
   f+HMrEtpHvvsA61bw223ZV2JJClDNsFSmRvKjYynF6/yo6xLaR4hJGeQO+ccR4MlqYzZBEtlrAWL
   +DV/L59R4Er9+yfbpt16a9aVSJIyYhMslbGfcxsz6Mz/+GnWpTSvEJKdIv74R1i0KOtqJEkZsAmW
   ylbkDP6aGwWudQFtadp9d0eDJamM2QRLZWonHqMN87mXAVmXko3KucGOBktSWbIJlsrUqfyLCziF
   WM5PA7vtBiuvDLfcknUlkqRm5skypDJR9WQZGzKBsfSlG1P4juXrm0TRniyjJg8/DCedBG+8AS1b
   plCLJClrnixDUo1O5kIu4xcNaIBL0C67wCqrwOjRWVciSWpGjgRLZaJyJLgTM5lEDzbkbaazWkOS
   KKmRYIBHHoETToA333Q0WJJKgCPBkpZyHJdzB/s2sAEuUTvvDJ07w6hRWVciSWomjgRLZSKEQBu+
   Ywrd2JWHeZNNGppEyY0EAzz6KPzyl/DWW44GS1KRcyRY0hIOYhTj6dWIBriE7bQTrLoq3HZb1pVI
   kpqBTbBURk7gYi7k5KzLKEwhwO9+B3/6E1RUZF2NJKmJ2QRLZWJLoBOzeIB+WZdSuPr1gzZt4O67
   s65EktTEbIKlMnE8cBm/oILSnO8aQmj8pUULOOssOO88cA2DJJU0m2CpHMycyV7AtRyRdSVNKKZw
   AQYOhG++SU6iIUkqWTbBUjm45hrGAJ/TKetKCl+LFvDb3yZzgyVJJcsmWCp1ixbBZZdxSdZ1FJMD
   D4Rp0+B//8u6EklSE7EJlkrdfffBaqvxUtZ1FJNWreCMM5K5wZKkkmQTLJW6Sy6B44/PuoriM3Qo
   TJgA48ZlXYkkqQl4xjiplE2aBNttBx9+SFh+edI6O1shnjGuSc4895//wCOPwJ13ppAtSWounjFO
   KneXXQZHHAHLLZd1JcXpqKOSkeDXX8+6EklSyhwJlkrV3Lmw1lrw4ovQrRshFNqoa5pZTTQSDPD3
   v8PLL8OoUSnkS5KaQz4jwTbBUoFLmtf6OwQ4CBiwxLWF1HCmmdWETfBXX8E668ALL0D37ikcQ5LU
   1JwOIZWM+p/44Wi25yruYIkTQaj+2reHY46B88/PuhJJUoocCZYKXEOmMWzIBB5jJ9biQxbSujKp
   3jnLqCilnDSzmnAkGODTT6FnT3jnHejcOYXjSJKakiPBUpk6iqu5nsOqNMBqlNVXh/33T3aLkCSV
   BEeCpQJX35HgNszjI7qyLc/wLutVTapXTi0VpZSTZlYTjwRDMgq83Xbw/vvQrl0Kx5IkNRVHgqUy
   tDd38gabVGuA1Wjrrw99+sC112ZdiSQpBTbBUok5mqu4iqOzLqM0/frXyQK5BQuyrkSS1Eg2wVIJ
   WYf32JxX+S/7ZF1KaerdO9km7ZZbsq5EktRINsFSCTmSa7iZg5mHZ4hrMr/+dXICDdc4SFJRswmW
   SkRLFnI413E1R2VdSmnr1y95++CD2dYhSWqUVlkXICkde3AfU+jGW2ycdSlFLZ8z9A0Gjurfn51q
   uY+74UhSYXMkWCoRR3G1o8CpqPtsfLcwn3VYmy15fhn3kSQVOvcJlgpcPvsE/4DPmMgGrMlUvmFZ
   e9gW2p68aWY1f87J/JuteY5BjKoxx+c7ScqO+wRLZWIwI7iTvWtpgJW2azmCXXmYrnyYdSmSpAaw
   CZZKwGFcz/UclnUZZWUO7bmBQzmBi7MuRZLUAE6HkApcXdMhNuNVxjCQ7rxHrPX/2kKbepBmVjY5
   3XifF9iSbkypNgrvdAhJypLTIaQycCg3cCND62iA1RSmsA5P0IfDuD7rUiRJ9VTnX80QQr8Qwtsh
   hEkhhN/UcPveIYTXQgivhBBeCiHUtmuQpBS1YgGDGcGNDM26lLJ1AadwMhcSqMi6FElSPdTaBIcQ
   WgIXA/2AjYBBIYSe1e72SIxxsxjjj4DDgCubolBJS+vP/UyiB++yXtallK2n2Y7ZdGAA92ZdiiSp
   HuoaCe4NTI4xTokxLgBGAXtXvUOM8ZsqH7YDZqZboqRlcUFcIQhcwCmcwgVZFyJJqoe6muAuwEdV
   Pp6au24JIYSBIYQJwP3ASemVJ2lZOjGTnXiMW9k/61LK3q3sz/q8w2a8mnUpkqQ81XXa5LyWN8cY
   xwBjQgg7ADcBG9R0v2HDhi1+v2/fvvTt2zevIiUtbRAjuZcBfEWHrEspewtowyUcz//xbw53kZwk
   NbuxY8cyduzYen1OrVukhRC2BobFGPvlPj4TqIgx/q2Wz3kX6B1jnFXterdIkxpgWVukvcBP+C1/
   5mF2yzepxpwGVJRSTppZ2eeswiwmsx4b8jbTWd0t0iQpQ2lskfYi0COE0C2E0AY4ELir2kHWDclf
   aUIIWwBUb4AlpWtj3mB1PuVRds66FOV8TifuYF+O5JqsS5Ek5aHW6RAxxoUhhBOAB4GWwDUxxgkh
   hGNzt18B7AcMDSEsAL4GDmrimqWydyg3cBOHUEHLrEtRFZdwPGMYyN+zLkSSVCfPGCcVuOrTIVqy
   kI/oyo48zkQ2rE8SWU8ZaLqswsl5hm34G88xxuc7ScqMZ4yTStBuPMQHrF3PBljN5RKO5/isi5Ak
   1ckmWCoyQxjOTRySdRlahlvZn00BJk7MuhRJUi1sgqUisgLfMIB7uYUDsi5FyzCftsnSuEsvzboU
   SVItbIKlIrIXd/Es2zCTzlmXolpcDnDzzfD111mXIklaBptgqYgMZgQjGJx1GarDRwA77AAjRmRd
   iiRpGdwdQipwlbtDrMIs3qM7azKVr1mpIUkUyg4K6WcVXk586CE4/XR49VUItS5QliSlzN0hpBKy
   H7fzAP0a2ACr2e28M3z3HTz9dNaVSJJqYBMsFYnBjGAkg7IuQ/lq0QJ++UsXyElSgXI6hFTgQgis
   yYe8xmb8kE+YT9uGJlFYUwbSzCq8nBgjfPEFrLMOTJ4Mq66aQq4kKR9Oh5BKxIGM5g72bUQDrEx0
   7Ah77gk33ZR1JZKkamyCpSLgrhBF7Oij4aqrwFfCJKmg2ARLBW4DYHU+5Qn6ZF2KGmKHHaCiAp55
   JutKJElVtMq6AEm1GwSM4iAqaJl1KaqHUGVbtNOATbbfnsMbmOV6CklKnyPBUiGLkcHgrhBFKS6+
   3MB0BtKBDnyxxPX5XSRJTcEmWCpkL76YvOEnGReixphJZx5kd4YwPOtSJEk5NsFSIRsxguTEu55x
   rNhdxdEczVU4uitJhcEmWCpUixbB6NGMzLoOpeIxdqI9X/ETXsy6FEkSNsFS4XriCfjhD5mYdR1K
   RaQFV3NUbjRYkpQ1m2CpUI0YAYPdG7iUXMfh7M+ttGNO1qVIUtmzCZYK0bx5cMcdcOCBWVeiFH3K
   D3mCPhzALVmXIkllzyZYKkT33w+bbgprrpl1JUrZ9RzGodyQdRmSVPZsgqVC5FSIknUvA9iAiazL
   5KxLkaSyZhMsFZo5c+DBB2G//bKuRE1gIa0ZwWCGcmPWpUhSWbMJlgrNmDHQpw906pR1JWoilVMi
   AhVZlyJJZcsmWCo0ToUoea+zGZ+zCjvyeNalSFLZsgmWCsn06fDss7DnnllXoiZ2PYdxGNdnXYYk
   lS2bYKmQ3HorDBgAK66YdSVqYiMYzJ7czUp8lXUpklSWbIKlQjJypFMhysRMOvMYO7E/t2ZdiiSV
   JZtgqVBMmQITJ8Juu2VdiZqJUyIkKTs2wVKhGDUKfv5zaN0660rUTO6nP+vzjnsGS1IGbIKlQuGu
   EGVnIa0ZzhDPICdJGbAJlgrB+PHwxRew3XZZV6Jm5p7BkpQNm2CpEIwcCYMGQQt/JcvNeDblCzqy
   A09mXYoklRX/4kpZi9FdIcrczRzMwdycdRmSVFZsgqWsPfccLL88bLZZ1pUoIyMZxL7cQVu+y7oU
   SSobNsFS1ioXxIWQdSXKyMesySv8iAHcm3UpklQ2bIKlLC1cCLfckswHVlm7mYM5hJuyLkOSyoZN
   sJSlRx+Fbt1g3XWzrkQZu4N92ZHHWYVZWZciSWXBJljKkgvilDOH9txPf0+jLEnNxCZYysq338Kd
   d8IBB2RdiQqEu0RIUvMJMcbmOVAIsbmOJRWCUMdCt/2AY4Hd8kpL43cnFFhOmlmlkdOKBXxMF7bm
   Od6n++IsnzslqX5CCMQYa/1D7Eiw1KTiMi+D2YcRXFvrfdJrNlUMFtKa0RzIEIZnXYoklTxHgqUm
   kowE1/wz34Ev+YC1WYsP+YoOdSUtM6eeFRVYTppZpZPTm+e5kaFsyNu5HEeCJam+HAmWCtS+3MGj
   7JxHA6xyM47etKCCn/Bi1qVIUkmzCZYyMIiRjMBdIVST4AI5SWoGToeQmsiypkOszie8xUaswTS+
   Y/l8kmrMaUBFBZaTZlZp5azLZJ5mO9ZkKgtp43QISaonp0NIBegAbuFO9s6zAVY5epf1eI/u7MrD
   WZciSSVXONCnAAAgAElEQVTLJlhqZoMZ4VQI1ckpEZLUtJwOITWRmqZDrMtknmJ71mQqi2iVb9JS
   OQ2sqMBy0swqvZxVmcEkerAms/na505JqhenQ0gF5iBGcSv716MBVrmaSWeeZjsGZl2IJJUom2Cp
   2USGMNypEMrbcIYwJOsiJKlE2QRLzeRHvEIb5vMcW2ddiorEnezNNgDTp2ddiiSVHJtgqZl8Pwpc
   6xQlabG5rMg9AKNHZ12KJJUcm2CpGbRgkSfIUIMMBxg+POsyJKnk2ARLzaAPT/Apq/M2PbMuRUXm
   YYD334fJk7MuRZJKik2w1AyGMJzhLnFSAywCOPBAGDEi61IkqaS4T7DURCr3CW7Ld0xjDXoxnml0
   aUgShbJ3bbo5aWaVak6SFZ97DoYOhbffhuCcckmqi/sESwVgAPfyCj9qYAMsAb17w6JF8NJLWVci
   SSXDJlhqYu4NrEYLAYYMcYGcJKXI6RBSEwkhsDKfM4VurM0HzGblhiZRWC/ROx2i+XKSrBgjvPMO
   9OkDU6dCy5YpZUtSaXI6hJSxfbmDR9ilEQ2wlLP++rDmmvDYY1lXIkklwSZYakLuCqFUOSVCklLj
   dAipiXQJgfF0ZA2mMY/lGpFUaC/ROx2i+XKSrMXPnZ9+Cj17wrRpsPzyKeVLUulxOoSUoUHAf9mn
   kQ2wVMXqq8OWW8Ldd2ddiSQVPZtgqYkMAadCKH1OiZCkVDgdQmoKb73FxxtvzFospILGruQvtJfo
   nQ7RfDlJ1hLPnV99BWutBe+9B6usktIxJKm0OB1CysqIEYyEFBpgqZr27WH33eHWW7OuRJKKmk2w
   lLYYYcQIfMFaTcYpEZLUaK3yuVMIoR/wb6AlcHWM8W/Vbh8C/JrkNcA5wC9ijK+nXKtUHJ59Ftq2
   5dWs61DJCGHJV/RaA9OAH4fAh/XIcUqaJH2vzpHgEEJL4GKgH7ARMCiE0LPa3d4Dfhpj3BQ4F7gy
   7UKlojF8eDJSJ6UmLnFZQOQ2jmUQf1nqtmVfJElV5TMdojcwOcY4Jca4ABgF7F31DjHGZ2OMs3Mf
   Pg+smW6ZUpGYNw9Gj4aDD866EpW44QxhiJNuJKnB8mmCuwAfVfl4au66ZTkSuK8xRUlF6557oFcv
   6NYt60pU4p5mO9rzFb1w5pkkNUQ+c4Lzfh0thLAjcASwXU23Dxs2bPH7ffv2pW/fvvlGS8Xhxhth
   6NCsq1AZiLRgBIMZwnDOYNOsy5GkTI0dO5axY8fW63Pq3Cc4hLA1MCzG2C/38ZlARQ2L4zYF7gD6
   xRgn15DjPsEqbTNmQI8e8NFHsNJKucVMhbTnbKHlpJlVqjm1Z23MG9zHHnRjCrHOF/aCC+MklY20
   9gl+EegRQugWQmgDHAjcVe1Aa5E0wAfX1ABLZWHECNhzT1hppawrUZl4k034go7swJNZlyJJRafO
   JjjGuBA4AXgQeAsYHWOcEEI4NoRwbO5ufwA6ApeFEF4JIYxrsoqlQnXDDXDooVlXoTLjAjlJahhP
   myylYfx46N8fPvgAWiZniXM6RHNmlWpO3Vlr8hGvsjlrMI35tK01x+dgSeXC0yZLzeXGG+GQQxY3
   wFJzmUpXxtOL/tyfdSmSVFRsgqXGWrgwOUGGUyGUEadESFL92QRLjfXww9C1K2y4YdaVqEzdxs/Z
   jYdoz+y67yxJAmyCpca74Qb3BlamvqQjj7ET+3JH1qVIUtGwCZYa48sv4f774aCDsq5EZc4pEZJU
   PzbBUmPceivssgt06pR1JSpz9/AztuBlfsi0rEuRpKJgEyw1hnsDq0DMYznGMJCDGJV1KZJUFGyC
   pYaaPBneeSfZH1gqAMMZwsHcnHUZklQUbIKlhrr+ehg8GFq3zroSCYCx9GU1PmNDJmRdiiQVPJtg
   qSEWLUqa4COPzLoSabEKWjKKg1wgJ0l5sAmWGuLBB6FLF+jVK+tKpCUMZwiDGUF6p22WpNJkEyw1
   xDXXOAqsgvQKP2IebdmGZ7MuRZIKmk2wVF+ffQaPPurewCpQwT2DJSkPNsFSfd10E+yzD7Rvn3Ul
   Uo1GMJj9uZVWLMi6FEkqWDbBUn3E6FQIFbz36c5k1mM3Hsq6FEkqWDbBUn0880zSCG+3XdaVSLVy
   SoQk1c4mWKqPa66BI46AELKuRKrVLRzAHtzHinyddSmSVJBsgqV8zZkD//0vDB2adSVSnWbSmafZ
   joGMyboUSSpINsFSvkaPhr59YfXVs65EyotTIiRp2WyCpXxdfbUL4lRU7mRvtuFZOjM961IkqeDY
   BEv5eP11mDoV+vXLuhIpb3NZkXv4GQcyOutSJKng2ARL+bjiCjj6aGjVKutKpHpxSoQk1SzE2Dzn
   lw8hxOY6lpSqr7+GtdaC8eOhS5e8Py2EAKTxM1+qOWlmlWpO47NaspCprEkfPmOiz8GSykQIgRhj
   rVs5ORIs1WXkSOjTp14NsFQoFtGKmzmYw7IuRJIKjE2wVJsY4bLL4Ljjsq5EarDrOJyhAIsWZV2K
   JBUMm2CpNi++CF9+CbvumnUlUoO9xcZMBXjI0yhLUiWbYKk2l18OxxwDLfxVUXG7DuC667IuQ5IK
   hgvjpGX58kvo1g0mToTVVqv3p7swrjmzSjUnvawOBL7s0AHefRc6dWp8WZJUwFwYJzXGTTcl+wI3
   oAGWCs1sgD32gBEjsi5FkgqCTbBUkxiTvYFdEKdScsQRTomQpBybYKkmTz8NCxcmW6NJpWKnnWDW
   LHj11awrkaTM2QRLNbn0Ujj2WAi1TieSikuLFnDooY4GSxIujJOW9sknsNFG8P77sPLKDY5xYVxz
   ZpVqTppZgRgjvPcebLUVfPwxtGmTQq4kFR4XxkkNceWVcOCBjWqApYLVvTtssgncfXfWlUhSpmyC
   parmz08WxJ1wQtaVSE3n8MPh2muzrkKSMmUTLFV1++2w4YbJSJlUqvbbD555BqZNy7oSScqMTbBU
   1X/+4yiwSt+KKyaN8E03ZV2JJGXGJliq9NJLMHUq7LVX1pVITe+II+Caa5I9sSWpDNkES5Uuvhh+
   +Uto1SrrSqSmt802ye4QTzyRdSWSlAm3SJMAZsyAHj1g8mRYddVUIt0irTmzSjUnzazcFmlVXXQR
   PPssjByZQr4kFY58tkizCZYA/vpXmDgx1ZMI2AQ3Z1ap5qSZVUMT/MUXsM46MGkSdO6cwjEkqTC4
   T7CUj4ULkzPEnXhi1pVIzatjR9h7b7jxxqwrkaRmZxMsjRkDXbvCFltkXYnU/I45JjlBjK/USSoz
   NsHSBRfAqadmXYWUjW23TRaDukBOUpmxCVZ5e+655IQBAwdmXYmUjRDg2GOT0WBJKiMujFN5O/DA
   ZKuo//u/1KNdGNecWaWak2ZWDQvjKn3+OXTvnuruKJKUJRfGSbX54AN45JHkpAFSOVtlleQkMTfc
   kHUlktRsHAlWSUhGXevnHyRjbKenXk1VhTSqWGg5aWaVak6aWbX/jmwLXAtsmEeSz+WSCp0jwSoz
   Me9LO77icFbhP7xf7bb65dR+kQrNsn9en6GChWxEHx6v9X6SVCpsglWWjuBaHmMnPqBb1qVIBSJw
   Gb/gl1yadSGS1CycDqGSUJ9FaC1YxCR6MIThPMc21ZPyzsmjqpSySjUnzaxSzUkzq+6clfiKKXSj
   F+OZRpdl5vhcLqnQOR1CqsHe3Ml0flBDAyyVtzm0ZySDOJYrsi5FkpqcI8EqCfUZCX6S7fkPJ3IL
   B9aUlHdOHlWllFWqOWlmlWpOmln55fTkLR5jJ9bmA+bTtsYcn8slFTpHgqVqtuVp1mAat7Nf1qVI
   BWkCG/EmG7Mft2ddiiQ1KZtglZVf8Q/O5zQW0SrrUqSCdTEncAIXZ12GJDUpp0OoJOQzHWJDJjCW
   vqzD+3zLCstKqjOnHlWllFWqOWlmlWpOmln557RkIe/RnX34Ly/z46VyfC6XVOicDiFVcTr/5BKO
   r6UBlgSwiFZcxi84nkuyLkWSmowjwSoJdY0Er8HHjKcXPZjE53SqLanWnHpWlVJWqeakmVWqOWlm
   1S9nVWbwDuuzHpOr/c44Eiyp8DkSLOWczIXcxCF1NMCSKs2kM3exF0dwbdalSFKTcCRYJaG2keD2
   zOY9urMFL/Mha9eVtMycBlSVUlap5qSZVao5aWbVP+cnvMAtHMB6TKaClotzfC6XVOgcCZaA47ic
   ++mfRwMsqaoX2ZJP+CH78N+sS5Gk1DkSrJKwrJHgNszjfdahHw8wnk3zSaoxp4FVpZRVqjlpZpVq
   TppZDcvZj9s4lX+xHc8szvG5XFKhcyRYZW8oN/Iam+XZAEuq7r/sw+p8ytY8m3UpkpQqm2CVrJYs
   5Az+ynmclXUpUtGqoCX/5v84jfOzLkWSUmUTrJJ1IKOZypo8zfZZlyIVtWs5gr6MZR3ey7oUSUqN
   TbBKUqCCsziPP/G7rEuRit43tONqjuJkLsy6FElKjU2wStI+/Jc5rMQj7JJ1KVJJ+A8ncgg3sXLW
   hUhSSmyCVYJilVHgWheGSsrTNLpwDz/jmKwLkaSU1NkEhxD6hRDeDiFMCiH8pobbNwwhPBtC+C6E
   cFrTlCnlrz/305JF3MPPsi5FKin/4lROBJg/P+tSJKnRam2CQwgtgYuBfsBGwKAQQs9qd5sFnAj8
   s0kqlOol8nvOze0I4SiwlKbX2JyJAKNHZ12KJDVaXSPBvYHJMcYpMcYFwChg76p3iDHOiDG+CCxo
   ohqlvO3I43TkC25nv6xLkUrSPwD+/neoqMi6FElqlLqa4C7AR1U+npq7TipIZ3Eef+FMKmiZdSlS
   SXoQoE0buOeerEuRpEapqwn23JgqGjvwP9bhfUYwOOtSpNL229/CeeeBp0+WVMRa1XH7x0DXKh93
   JRkNbpBhw4Ytfr9v37707du3oVHSUs7hbP7IH1hI66xLkUrbPvvA734Hjz0GO++cdTWSxNixYxk7
   dmy9PifEWv6TDyG0AiYCOwPTgHHAoBjjhBruOwyYE2Os8dyaIYRY27GkxtgxBK5kPXoygUV1/m9X
   m0B6L4CklVWqOWlmlWpOmlnp5cQY4YYbkstjj6WQKUnpCiEQY6x1hXytTXAupD/wb6AlcE2M8S8h
   hGMBYoxXhBBWB14A2gMVwBxgoxjj19VybILVNGLkyRYtuIKbGM7BjQwr7ealsHLSzCrVnDSzUm6C
   FyyAHj1g5EjYZpsUciUpPak0wSkWYxOspvHww7y9225szMIUFsSVdvNSWDlpZpVqTppZKTfBAJde
   CvffD3ffnUKuJKUnnybYM8apuMUIf/gDw8AdIaTmdsQR8PLLyUWSioxNsIrbAw/AnDncmnUdUjla
   bjk44wyosuhZkoqFTbCKV24UmGHDcNt+KSNHHw2vvALjxmVdiSTVi02witc998D8+bDvvllXIpWv
   5ZZL9g0+++ysK5GkerEJVnFatAjOOgv++Edo4Y+xlKkjj4QJE+CZZ7KuRJLyZveg4jR8OKy0Euy1
   V9aVSGrTJjl5hqPBkoqIW6Sp+Hz3HWywQdIIb789kGyFUljbSKWZVao5aWaVak6aWU2wRVpVCxbA
   hhvCddfBT3+awnEkqeHcIk2l6dJLYfPNFzfAkgpA69bJLhFnnJEsWpWkAmcTrOLy5Zfw17/Cn/+c
   dSWSqhsyBObOhTvuyLoSSaqTTbCKy9/+BnvuCRtvnHUlkqpr0QL+8Y9kNHjBgqyrkaRaOSdYxePj
   j6FXL3jtNejadYmbnBNcjDlpZpVqTppZTTwnuKrdd08WrR5/fArHk6T6y2dOsE2wisfRR8MqqySj
   wdXYBBdjTppZpZqTZlYzNsGvvZY0wu+8A+3bp3BMSaofm2CVjrfegj59kj+qHTsudbNNcDHmpJlV
   qjlpZjVjEwxw2GHJKzbnnpvCMSWpfmyCVRpiTEaVBgyAk0+u8S42wcWYk2ZWqeakmdXMTfBHHyW7
   uLz8Mqy9dgrHlaT8uUWaSsM99yR/UH/5y6wrkZSvrl2Tf1pPOy3rSiSpRjbBKmzz5sGpp8IFFyT7
   kEoqHr/6VTIS/PDDWVciSUuxCVZhu+ii5Oxw/fplXYmk+lp+efj3v+HEE2H+/KyrkaQlOCdYheuz
   z5L9gJ95BtZfv9a7Oie4GHPSzCrVnDSzmnlOcKUYk/n8O+6YjAxLUjNwYZyK2xFHJDtBnH9+nXe1
   CS7GnDSzSjUnzayMmmCASZNg663h9dehS5cUapCk2tkEq3g9+SQMGpRsjZbHPqM2wcWYk2ZWqeak
   mZVhEwzwu98lv8+33w6h1r9LktRo7g6h4jR/Phx3XDKX0I32pdLwu9/BhAlw221ZVyJJgE2wCtH5
   5yf7iu63X9aVSErLcsvBNdfASSfBrFlZVyNJTodQgXnvPdhyS3jxRVhnnbw/zekQxZiTZlap5qSZ
   lfF0iEonnwxffAE33phCLZJUM6dDqLjECCecAKefXq8GWFIR+fOf4amn4L77sq5EUpmzCVbhGD4c
   PvzQM0xJpWzFFeGqq+DYY+Hzz7OuRlIZczqECsO0abD55nD//fDjH9f7050OUYw5aWaVak6aWQUy
   HaLSKafABx+4W4SkJuF0CBWHGOGYY5IdIRrQAEsqQn/9K7z/Plx5ZdaVSCpTrbIuQOL662HqVLjj
   jqwrkdRc2raFkSNhhx1g++2Ts0NKUjNyOoSy9dFHsMUW8MgjsNlmDY5xOkQx5qSZVao5aWYV2HSI
   SldfDRddBOPGJduoSVIKPGOcCtuiRbDrrrDTTslG+o1gE1yMOWlmlWpOmlkF2gTHCAcdlCyYu+Ya
   5wdLSoVzglXYzjsv+QN45plZVyIpKyEkze+LL8LFF2ddjaQy4pxgZWPsWLjsMnjpJWjZMutqJGWp
   XTsYMwa23RY22QR23DHriiSVAUeC1fxmzICDD04WxK2xRtbVSCoE3bsne4UPGpTsGiFJTcwmWM2r
   ogKGDk2a4N13z7oaSYVk552T6VF77ZWcWlmSmpBNsJrXWWfB3Llw7rlZVyKpEJ10EuyyC+y5Z/Jc
   IUlNxN0h1HyGD4ff/z7ZCmnVVYHKXR3SUkgr6NPMKtWcNLNKNSfNrDRz0rHMvwmVrxjNnp3sH966
   dWrHlFQe3CJNhWPcOPjZz+DRR6FXr8VXF97WZqXevBRSTppZpZqTZlbh5dT6N2HBAth7b/jBD+Da
   a6GFL1xKyp9bpKkwfPwx7Ltvsil+lQZYkpapdWu49VaYNCk5pXpFRdYVSSoxNsFqWp9/niyAO/nk
   ZLGLJOVrxRXhgQfgnXfg0ENh4cKsK5JUQmyC1XS+/hr22AP694fTT8+6GknFaKWV4L77YPr05Mxy
   8+dnXZGkEmETrKYxbx4MHJhsfP/3v3sqVEkNt8IKcNddyUjwXnslC+YkqZFsgpW++fOTDe87doQr
   rrABltR4bdsmc4S7d0/OLOcJNSQ1kk2w0vXtt8kiuEWL4OabPSWypPS0bg2XXJIslNt2W3jqqawr
   klTEbIKVnjlzYMAAaN8ebrstGbmRpDSFACeeCNddl/zDfdFF4PabkhrAfYLLSO/eP+W77xqf06oV
   XHfdf9hss82+v/Lzz5NFcL16weWX5z0C7D7B5ZyTZlap5qSZVXg5jf6b8O67yWK5H/4w2Us4dxIe
   Scpnn+BWzVWMsvfCC08CY2nsGZ9WXPEEZlddmPLOO8kpTvfay0VwkprPuuvC00/z97ZtGdS5M4cB
   jzUy0sEaqXzYBJedn9LYJrhVq5W//+CRR2DIEDjvPDjqqMaVJkn11aYNvwEe5QGu42geYRd+xT/4
   nE4NCPMfeKmcOCdYDRNjskDl4INh9GgbYEmZeojd2Zg3mcNKvMEmDGY46U1JkVSKnBNcRpL5txU0
   drRjrfbb8uIWben8xRdw++3JS5KNqqmw5ik6l7O5ctLMKtWcNLMKLyetvwnVn0d68zyXcxzfsjyn
   80+eZdtmr0lStvKZE+xIsOqlN8/zxJxXmN+xIzz3XKMaYElqCuPYip/wIpdzHKM4iNvYjw14O+uy
   JBUYm2DlpTXz+T1/5G725PfLr8O7//d/sNxyWZclSTWqoCU3MZQNmMg4evM/fspoDmBzXsm6NEkF
   wiZYddqCl3iBLdmK5/kxL3F3a7chklQcvmN5/s5v6M57PMfW3MPPuJc96Mf9BCqyLk9ShmyCtUwr
   8jV/4QzuYw/+wa/4Gfcwla5ZlyVJ9fYN7biAU1mXd7md/TiPs5hED07jn6zKjKzLk5QBm2DVIHIA
   o5lAT7rwMZvxGsM5GLcPklTs5rEc13IkP+YlhjCcTXmdSfTgHgYwGODrr7MuUVIzcXeIMpLP7hCb
   8wr/4lRW4XNO4GKeYoel7tOhw0+ZPfvJFCsrrBXrrupvrpw0s0o1J82sQsxJU8NrWpGv2Zs7GczB
   DOjQAfr3h733ht13h44d650XUjxhkH83pYZxdwjlrTvvMpzB3Mce3MIB/JiXamyAlxRTuEgqX2k8
   hzT+eeQb2jGCIfwMYNIk6NMHbr4Z1l47ef8f/4Dx46GiPnOIs/+6JNXOkeAyUtNIcGem83vOZRAj
   uZCTuYBT+IZ2teZ8PxJcaKNKhZSTZlap5qSZVao5aWaVak6aWdX2CZ47Fx5/HO69Fx5+GL74AnbY
   IWmM+/SBTTeFli2XTklx/3P/bkoNk89IsKdNLlOr8wmn808O5zpuZCg9mcBMOmddliQVjhVWgAED
   kgvAxx/DE08kl8svTz7efHPYcsvk0rs3dO+ebc2S8uZIcBkJIdCVKfyafzCYEdzAofyT05lGl3rl
   OBLc3FmlmpNmVqnmpJlVqjlpZtVz5PXLL+HFF+GFF76/fPMNY7/4gtc5kfH04nU25Q02YS4rNn09
   khZzJFjfmzyZK4F92YKrOJqeTGA6q2VdlSQVr5VXhl12SS6VPvuM81ZfnV50Yzue5hdcxoa8zcd0
   WdwUV17eozvRpTlSZmyCS1mM8L//wQUXwNNP8wmwPhP5HE92IUlNYrXVeAR4hFMXX9WShfRg0uL2
   93CuY1NeZxU+5w02WdwUv8ZmjKcXX9Ehu/qlMuJ0iFI0fz7cckvS/H79NZxyCgwdSlhxReraIi0f
   Todo7qxSzUkzq1Rz0swq1Zw0s9KZfpDvwrgOfEkvxrMpr7MZr7Epr7MJbzCTVXNN8d38/pZbkgV4
   661X4yI8STXLZzqETXApmToVrr0WrrwSNtgATj012e+yRfJyWz77BOfDJri5s0o1J82sUs1JM6tU
   c9LMat4muOYKKliXd3Njwz/n7IED4bXXYPp02GSTZCFe5aVXL1ixIXONpdJnE1wOFi2CBx+EK66A
   J5+EAw+E446DzTZb6q42wc2Zk2ZWqeakmVWqOWlmlWpOmllpnsAj3XpWAjYFNq9y2Qj4EHi12uWz
   mqpJ6e9voZ0IpNDqUeFwYVwpmzIFbroJrr4aVlsNjjkGhg+HdrXv8StJqk1hNtNzgKdzl0qtWMAG
   TGRzXmVzXuX03NtFtMy9V3kZnAyYpDadojAfo8ZJ+wyGKgaOBBeTWbPg1luTZnfCBDjgADjqKNhi
   i7w+3ZHg5sxJM6tUc9LMKtWcNLNKNSfNrFLIiXTh48Xt72a8xubcRo8VVoCNN4aePWH99b+/rLde
   vaZUFNqJQAqtHhUOp0OUgunT4e67YcyYZKeHfv3g4IOTc9q3aVOvKJvg5sxJM6tUc9LMKtWcNLNK
   NSfNrNLNibNnJ6d+fuedJS+TJ0OnTtCjR3Ka6K5dk8taa33/fvv23ycVWNNZaPWocNgEF6MY4a23
   4IEHksZ3/HjYbTcYODA5a1GHhm+dYxPcnDlpZpVqTppZpZqTZlap5qSZVbo5y/z7W1EBH30EkybB
   hx8m73/00ZLvt2iRTLvr3Jk7n32W6RzFDDoznR8wg87MoDOz6cBsOvAlKzObDnzHctT+t8YmWE0r
   lTnBIYR+wL+BlsDVMca/1XCfi4D+wFzgsBjjKw0ruQzFmDzZPPYYPPpocll+edh1V/jtb2GnnaBt
   26yrlCSVohYtkhHgtdeu+fYYYfZsmDEDpk/nuu235wdsSWdm0I0p9GYcnZmRa4FnszJf0oHZtKBi
   qcZ4Nh34mnZ8y/LMheRv3AorJH/zVljh+/fbtk3mL7dsCa1aLfm26vsh0AuIjKeCFlTQgkhY4m3+
   1wFz5ybHT3GxnQpbrU1wCKElcDGwC/Ax8EII4a4Y44Qq99kDWC/G2COEsBVwGbB1E9Zc3D7/HF5+
   GcaNg+efT95WVMCOO8LOO8Mf/+i55yXGAn0zrkGFZyz+XDSzEJIz4628MvTowZ0AHFPnp7Vh3hJN
   ceVlRb5hBeayAiRzkefOTf4uzp0L336bvJ03L1nIt2gRLFy47Pdj5GbgJfZiK5YnEGlBxVJv87mu
   JcCqq8J33yV1tWv3/dt27ZJXYTt1Si6rrLLk28pL585JI6+iUet0iBDCNsDZMcZ+uY/PAIgx/rXK
   fS4HHo8xjs59/DbQJ8b4WbWs8pkOMWdOMrr74Yfw/vvJIrY330ymOXz7bbLx+VZbQe/eydu11mqW
   /zydDtGcOWlmlWpObVnDcpfG5qRVT1Y5aWaVQs4wav65KIWvrWlz0t0irXBqSuo5m/o9X9RSz6JF
   8M03yeXrr5PLnDnJaPisWUnDPmvW0u/PmpWMlrdokTTDq66a39uOHT0JShNJYzpEF+CjKh9PBbbK
   4z5rUvNWhcWjoiL5j3Tu3OSXofLtN98kvwwzZ35/qfzhnzo1aXy/+y5pbCtfYurZE/bcEzbaCLr8
   f3v3F1plHcdx/P1p7uAkZOQuIjfIiy7UKCqQKKIgjDEyL7wIIaIiCGol0l8TwvvoH11F/ygqdmER
   CZJ5UXQTZbUU0zKjgiXhKh0WjCn7dvF7Tjsez9HJ5vN79HxecDjP8+wZz5dxYJ/z+7vUXS1mZmZV
   1NWVJgI2TAactYiUEcbHZ/JB/Xh8PC1t2nxtYiK1sjcH5L6+VEO9Jbrda+HCNEm+Vvt/iIjN3plC
   8MZRXvkAAAPgSURBVGy/pjX/1avT5Ds2BuvWndyV0qp7pfF9cjJ1x/T0pO6QRYvSe/24tzd1ffT1
   pckCK1ak4/7+FH6XLKnsB3Hx4jXMtSV4cnLv/BRjZmZ2oZBmwumyZbP7nRMn4MiRk4NxPUAfPZoy
   TL1Fuvl17FjKKlNT6TU9PROIazXo7j75WEot1Y3vra7t2DGnSfjnkzMNh7ge2NIwHGITMN04Oa4Y
   DvFZRIwU522HQ5yD+s3MzMzMTjHX4RBfA1dIuhw4BNwJrG+65yNgGBgpQvPR5gA8m0LMzMzMzMpy
   2hAcESckDQM7SEukvR4R+yU9UPz8lYjYLmlI0kHgX+Dec161mZmZmdkclLZZhpmZmZlZVVxU9gMl
   PSppWtIlZT/bqknSs5L2S9ot6QNJnTEi304haVDSD5J+kvRk7nosP0kDkj6V9L2kvZIeyV2TVYek
   LkmjkrblrsWqQVKvpK1FrthXDNVtqdQQLGkAWA38VuZzrfI+AVZGxNXAAWBT5nosg4bNeQaBFcB6
   ScvzVmUVcBzYGBErSRsxPeTPhTXYAOyjSqtSWW4vAdsjYjlwFbC/3Y1ltwQ/DzxR8jOt4iJiZ0RM
   F6dfktaZts6zCjgYEb9GxHFgBFibuSbLLCL+iIjviuN/SP/QLstblVWBpH5gCHiNua79aReEoif5
   poh4A9LctoiYaHd/aSFY0lpgLCL2lPVMOy/dB2zPXYRl0WrjnaWZarEKKlYquob0ZdnsBeBx0lao
   ZgDLgHFJb0r6VtKrktruZX2mJdLOiqSdwKUtfrSZ1MV9W+Pt8/lsq7bTfDaejohtxT2bgamIeK/U
   4qwq3J1pbUm6GNgKbChahK2DSbodOBwRo5JuyV2PVcYC4FpgOCJ2SXoReAp4pt3N8yYiVre6LulK
   Ujrfnfb5ph/4RtKqiDg8nzVYNbX7bNRJuofUrXVrKQVZFf0ODDScD5Bag63DSeoG3gfeiYgPc9dj
   lXADcIekIWAhsFjS2xFxd+a6LK8x0qiDXcX5VlIIbinLEmmSfgGui4i/S3+4VY6kQeA50k6Df+au
   x/KQtAD4kfRF6BDwFbA+ItpOarALn1LLyVvAXxGxMXc9Vj2SbgYei4g1uWux/CR9DtwfEQckbQF6
   IqLlakPz2hJ8FtztaY1eBmrAzqKn4IuIeDBvSVa2dpvzZC7L8rsRuAvYI2m0uLYpIj7OWJNVj3OF
   1T0MvCupBvzMaTZx82YZZmZmZtZxSt8sw8zMzMwsN4dgMzMzM+s4DsFmZmZm1nEcgs3MzMys4zgE
   m5mZmVnHcQg2MzMzs47jEGxmZmZmHcch2MzMzMw6zn8reaMqzYRoFgAAAABJRU5ErkJggg==
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>
   <div class="cell border-box-sizing text_cell rendered">
   <div class="prompt input_prompt">
   </div>
   <div class="inner_cell">
   <div class="text_cell_render border-box-sizing rendered_html">
   <h4 id="QQ-plot-of-deviance-residuals">QQ-plot of deviance residuals<a class="anchor-link" href="#QQ-plot-of-deviance-residuals">&#182;</a></h4>
   </div>
   </div>
   </div>
   
   <div class="cell border-box-sizing code_cell rendered">
   <div class="input">
   <div class="prompt input_prompt">
   In&nbsp;[43]:
   </div>
   <div class="inner_cell">
       <div class="input_area">
   <div class="highlight"><pre><span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">(</span><span class="n">figsize</span><span class="o">=</span><span class="p">(</span><span class="mi">12</span><span class="p">,</span><span class="mi">8</span><span class="p">))</span>
   <span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">)</span>
   <span class="n">fig</span> <span class="o">=</span> <span class="n">sm</span><span class="o">.</span><span class="n">graphics</span><span class="o">.</span><span class="n">qqplot</span><span class="p">(</span><span class="n">resid</span><span class="p">,</span> <span class="n">line</span><span class="o">=</span><span class="s">&#39;r&#39;</span><span class="p">,</span> <span class="n">ax</span><span class="o">=</span><span class="n">ax</span><span class="p">)</span>
   </pre></div>
   
   </div>
   </div>
   </div>
   
   <div class="output_wrapper">
   <div class="output">
   
   
   <div class="output_area"><div class="prompt"></div>
   
   
   <div class="output_png output_subarea ">
   <img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAtQAAAGyCAYAAADEaFMqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz
   AAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl41eWZ//H3EwiETQSXsioYF0Rwqa3LaDVaBVrUui9g
   cKbttB1bcNqZ2irya+oybe10AaodW1srR8G1dUtlsRZRq9YVcNegKJsLoiySQJLn98c3hywkEDhJ
   Tpb367py5eR7vufkzoXGjw/3cz8hxogkSZKknZOT7QIkSZKktsxALUmSJGXAQC1JkiRlwEAtSZIk
   ZcBALUmSJGXAQC1JkiRlIGuBOoQwOITw9xDCSyGEF0MIk6qu9w0hzAshvB5CmBtC2DVbNUqSJEnb
   E7I1hzqE0A/oF2N8IYTQE3gWOB34N+DDGOO1IYQfAH1ijD/MSpGSJEnSdmRthTrGuCrG+ELV4/XA
   K8BA4DTg5qrbbiYJ2ZIkSVKr1Cp6qEMIQ4DDgKeAz8QY36t66j3gM1kqS5IkSdqurAfqqnaPu4FL
   Yozraj4Xk34Uz0aXJElSq9U5m988hJBLEqZTMcZ7qi6/F0LoF2NcFULoD7xfz+sM2ZIkSWoRMcaw
   reezOeUjAH8AXo4x/rrGU/cBF1U9vgi4p+5rAWKMfrTRjx/96EdZr8EP//w64od/dm37wz+/tvvh
   n13b/miMbK5QHwNcCCwKITxfde0y4KfAHSGErwFvA+dmpzxJkiRp+7IWqGOMj9HwCvlJLVmLJEmS
   tLOyvilRHU9BQUG2S1AG/PNru/yza9v882u7/LNr/7J2sEsmQgixLdYtSZKktiWEQGytmxIlSZKk
   9sBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRl
   wEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXA
   QC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBA
   LUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAt
   SZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGXAQC1JkiRlwEAtSZIkZcBALUmSJGWgc7YL
   kCRJkupTXLyAadPmUlbWma5dy5k0aRRjxx6X7bK2YqCWJElSq1NcvIBLLplDSck1W66VlEwGaHWh
   2pYPSZIktTrTps2tFaYBSkquYfr0eVmqqGFZDdQhhD+GEN4LISyuca0ohLAshPB81ceYbNYoSZKk
   lldWVn8jRWlppxauZPuyvUJ9E1A3MEfglzHGw6o+ZmehLkmSJGVR167l9V7Py6to4Uq2L6uBOsb4
   KLCmnqdCS9ciSZKk1mPSpFHk50+udS0//3ImTjw5SxU1rLVuSpwYQpgAPAP8V4zx42wXJEmSpJaT
   3ng4ffoUSks7kZdXwcSJY1rdhkSAEGPMbgEhDAHujzGOrPp6T+CDqqevAvrHGL9W5zUx23VLkiSp
   /QshEGPcZvdEq1uhjjG+n34cQrgRuL+++4qKirY8LigooKCgoLlLkyRJUjs3f/585s+fv0OvaY0r
   1P1jjCurHn8X+HyMcVyd17hCLUmSpGbX6leoQwizgOOB3UMI7wI/AgpCCIeSTPt4C/hmFkuUJEmS
   tinrK9Q7wxVqSZIktYTGrFBnew61JEmS1KYZqCVJkqQMGKglSZKkDBioJUmSpAwYqCVJkqQMGKgl
   SZKkDBioJUmSpAwYqCVJkqQMGKglSZKkDBioJUmSpAwYqCVJkqQMGKglSZKkDBioJUmSpAwYqCVJ
   kqQMGKglSZKkDBioJUmSpAwYqCVJkqQMdM52AZIkSWo7iosXMG3aXMrKOtO1azmTJo1i7Njjsl1W
   VhmoJUmS1CjFxQu45JI5lJRcs+VaSclkgA4dqm35kCRJUqNMmza3VpgGKCm5hunT52WpotbBQC1J
   kqRGKSurv7mhtLRTC1fSuhioJUmS1Chdu5bXez0vr6KFK2ldDNSSJElqlEmTRpGfP7nWtfz8y5k4
   8eQsVdQ6hBhjtmvYYSGE2BbrliRJauuKixcwffo8Sks7kZdXwcSJJ7frDYkhBGKMYZv3tMVgaqCW
   JElSS2hMoLblQ5IkScqAgVqSJEnKgIFakiRJyoCBWpIkScqAgVqSJEnKgIFakiRJyoCBWpIkScqA
   gVqSJEnKgIFakiRJyoCBWpIkScqAgVqSJEnKgIFakiRJyoCBWpIkScqAgVqSJEnKgIFakiRJykDn
   bBcgSZKkllFcvIBp0+ZSVtaZrl3LmTRpFGPHHpftsto8A7UkSVIHUFy8gEsumUNJyTVbrpWUTAYw
   VGfIlg9JkqQOYNq0ubXCNEBJyTVMnz4vSxW1HwZqSZKkDqCsrP7GhNLSTi1cSftjoJYkSeoAunYt
   r/d6Xl5FC1fS/hioJUmSOoBJk0aRnz+51rX8/MuZOPHkLFXUfoQYY/a+eQh/BMYC78cYR1Zd6wvc
   DuwNvA2cG2P8uM7rYjbrliRJag12dGpHcfECpk+fR2lpJ/LyKpg48WQ3JG5HCIEYY9jmPVkO1F8A
   1gMzagTqa4EPY4zXhhB+APSJMf6wzusM1JIkqUOrb2pHfv5kpk4dbUhuQo0J1Flt+YgxPgqsqXP5
   NODmqsc3A6e3aFGSJEltgFM7Wo/W2EP9mRjje1WP3wM+k81iJEmSWiOndrQerTFQb1HV12FvhyRJ
   Uh1O7Wg9WuNJie+FEPrFGFeFEPoD79d3U1FR0ZbHBQUFFBQUtEx1kiRJrcCkSaMoKZlcp4f6ciZO
   HJPFqtq++fPnM3/+/B16TVY3JQKEEIYA99fZlLg6xvizEMIPgV3dlChJkjqSxk7vaPdTO8rK4JFH
   YNSorJXQFqZ8zAKOB3Yn6Zf+f8C9wB3AXjg2T5IkdTAdfnpHjPDEEzBjBtx5J4wcCQ88AD17ZqWc
   Vh+od5aBWpIktVejR1/B3LlX13N9CrNnX5WFilpISQmkUnDLLZCbC4WFMH487L13VstqTKBujT3U
   kiRJHVaHmt7x0Udwxx1JkH7jDTj/fLjtNjj8cAjbzLCtioFakiSpFWn30zs2bYK//jUJ0Q89BKNH
   w2WXJZ9zc7Nd3U5p1WPzJEmSOppJk0aRnz+51rVkesfJWaqoCcQITz4JF18MAwbAr34FX/oSLF2a
   rFCfckqbDdPgCrUkSVKrkt54OH36lBrTO8a0zQ2JS5YkPdG33JK0cBQWwjPPwJAh2a6sSbkpUZIk
   SU1nzZpkOkcqBa++CuedBxMmwOc/36b6otOc8iFJkqTmt2kTzJ6djLqbNy+ZG11YCGPGQJcu2a4u
   IwZqSZIkNY8Y4emnkxB9++0wbFgSos85B/r0yXZ1TcaxeZIkSWpab7+d9ESnUlBZmYTop56CffbJ
   dmVZY6CWJEnStn38Mdx1V7Ia/fLLcO65cPPNcOSRbbIvuqnZ8iFJkqStbd4Mc+YkIXrOHDjppGQ1
   +stfbvN90TvCHmpJkqRWqLh4AdOmzaWsrDNdu5YzadKo1jEWL8ZkrF0qlZxYuN9+SYg+91zo2zfb
   1WWFPdSSJEmtTHHxAi65ZA4lJddsuVZSkhzkkrVQvXQp3HprEqQ3bUpC9D/+Afvum5162hhXqCVJ
   kppBQ6vQo0dfwdy5V291/+jRU5g9+6qWK3Dt2uq+6MWLk1XowkI4+mj7omtwhVqSJCkLtrUKXVZW
   f/wqLe3U/IWVl8PcuUmInj0bTjgBJk2CsWOha9fm//7tlIFakiSpiU2bNrdWmAYoKbmG6dOn0LVr
   /X/LnpdX0TzFxAjPPVfdFz10aLISfd11sNtuzfM9OxgDtSRJUhPb1ir0979/IiUlk2sF7vz8y5k4
   cUzTFvHuu9V90Rs3woUXwqOPJhsN1aQM1JIkSU2sa9fyeq/n5VVs2Xg4ffoUSks7kZdXwcSJY5pm
   Q+K6dXD33UlLx8KFcPbZcMMNcMwx9kU3IzclSpIkNbH6eqjz8y9n6tQmCs41lZfDvHnJSvRf/wrH
   H5+0dJxyCuTlNe336oCcQy1JkpQlxcULmD59Xo1V6JObLkzHCC+8kIToWbNgr72SEH3eebDHHk3z
   PQQYqCVJktqX5cuTvugZM2D9+iREX3ghHHBAtitrtwzUkiRJzazZTz1cvx7+/OckRD/3HJx1VhKk
   jz0WcnKa7vuoXs6hliRJaiL1BWegeU49rKiAhx5KWjoeeAC+8AX4xjfg1FOhW7eMfg41PVeoJUmS
   tqP+TYaT2WWXNTz//PVb3b/Tpx4uXJiE6JkzYeBAmDABzj/fvugscoVakiSpCTR0UEufPhfVe/8O
   nXq4YkUSoGfMgE8+SXqiH34Yhg3LpGS1IAO1JEnSdjR0UAuU1Xt1u6cerl8Pf/lLshr99NNw5pkw
   bRocd5x90W2QgVqSJGk7GjqoZciQnvTt28hTDysqkpXnVAruuy85bOWrX4V77oHu3ZurdLUAe6gl
   SZK2Y1sHtQDbnje9eHESom+9Ffr1q+6L/sxnWvrH0E5wbJ4kSVIT2aGDWlauTPqiUylYvTrpiy4s
   hOHDW7ZoZcxALUmS1FI2bEjaN1IpeOopOP30JEQXFNgX3YY55UOSJGk7Gpov3ajDWioqYP78JETf
   cw8cfTRcdFFyEIt90R2GgVqSJHVY9fVGL1r0NaA3q1b9csu1rQ5reemlJETfcksyI3rCBPjpT5Me
   aXU4tnxIkqQOa/ToK5g79+o6V68A6l6D8wu+y6zT9kqC9Pvvw/jxSUvHiBEtUquyw5YPSZKkbah/
   vnT1tW58yle4l0JSHPfow7DXeXDttXDCCdBpBw5vUbtmoJYkSR1WffOlA5s5nr9TSIoz+Av/5Ahm
   MIEbCkZy780/y0KVau0M1JIkqcOouwHx6KMHUFKSHMxyIC9TSIoJOdNZk/N7bir/AZO5hlX0T2ZO
   f3dststXK2WgliRJHUJ9GxDXvPafTB+2nPwPBtC3dB3zB4yg5L+ms25oPi9Nn8cBpTdwSF4FEyeO
   aXjmtDq87W5KDCHsCyyLMZaGEE4ARgIzYowft0SBDdTkpkRJkrRdNVekX3zxFVavvp08NnIa91FI
   imN5jGf7D+aLN/8STjzRvmhtpUkOdgkhLAQOB4YAfwXuBQ6KMX65iercYQZqSZK0PTVXpAOVfIGv
   UkhnzuTPPMvhzGACf+EMPnf8/zJ/flG2y1Ur1VRTPipjjOUhhDOB6THG6SGE55umREmSpOYxbdpc
   OpcUcjWTGc+trONTZvB9RrKYFQzccl9eXkUWq1R70JhAvSmEMA6YAJxadS23+UqSJEnKwAcfwG23
   8esnbmJX/sBMxnE697CQT4C5UCNM5+dfzsSJY7JWqtqHxgTqrwLfBK6JMb4VQhgKpJq3LEmSpB1Q
   Wgr3358curJgAYwdS2q/47j2uRQVteJOYLfdzmfEiGHkudlQTaRRJyWGELoDe8UYX23+krbPHmpJ
   kkRlJTz+OMyYAXffDYcdxguHHEnRws18XNGDtWuXsXLlLrWOEM/Pv5ypUw3Rarym2pR4GvBzoGuM
   cUgI4TDgxzHG05qu1B1joJYkqQN7/fVkJfqWW6BHD175/LFcvaQLC1eXsWRJYOPG/9tya79+X2PA
   gDx69dqjakX6ZMO0dkhTBerngBOBv8cYD6u69mKMMWsH1xuoJUnqYD78EG6/PQnSb78NF1wAEyZQ
   vHwtl/zn3KrZ0lcAV2/10tGjpzB79lUtXbHaiaaa8rE5xvhxCLXepzKjyiRJkranrAweeABSKTbP
   e4jHdtmbB3c/lL/3H07FI2Xs8vy9W2ZLJ+qPNaWlzpZW82pMoH4phDAe6BxC2A+YBPyjecuSJEkd
   TXHxAqZNncN+H6zkSx++wAmrX+f1vN25q9sQfl95Lu+v+iOsWgDMAdKnHRbVeIfyet/XsXhqbo0J
   1BOByUAZMIvkn2L/3kSSJDWJ4uIF/O7S33DEa0u4vmINZXQlxbFM7HQgb228FdbUbOWYS3WYhtoh
   ehRJZKl+3rF4agnbDdQxxg3A5VUfLSaE8DawFqggaTs5oiW/vyRJaj7FxQu46Rf3csir/+Tk997g
   d5XruY2vcS4TeI7PAlOgIh2ia8aVutGlZohONht263Ye+fn9GTiwl2Px1CIaDNQhhPu38brYAlM+
   IlAQY/yomb+PJElqKWVlPHPlz8mdmuIPG1bxIIO4ij8wlycpr/UX4DUjSnkDjyEdomvPlv62IVot
   alsr1L9osSoats0dlZIkqQ2IEZ58kqVX/ZRdH5pHaezJ7eU/4RzOZi2/AsYCT9R5UUOtHPW1dcxm
   6tSLDdHKmgYDdYxxfgvWUW8JwEMhhArghhjj77NcjyRJ2hElJcms6FtuYX3pJu76dCjTN7/MUv4E
   fK3qpnRwrhuUR9G587coL/8/6rZy5OauJ4Rv15gtbVuHsmtbLR93xhjPCSG8SBJua4oxxoObtzSO
   iTGuDCHsAcwLIbwaY3w0/WRRUdGWGwsKCigoKGjmciRJ0natWQN33JHMi379dd4+8lh+sdsXuPW1
   Daz5OD3eblsbCaeQl7eU4cN7ceqpB/Pkk1MoLe1kK4dazPz585k/f/4OvabBg11CCANijCtCCHuz
   detFjDEu3akqd0II4UfA+hjjL6q+9mAXSZJai02b4MEHkyPAH3oIRo/m6WGH8e3732Dxq10pLf0t
   yXi7oqoX1B19t4Bu3a6rsZHQ0wzVemR0sEuMcUXVw4tjjD+o88Y/A36w9auaRgihO9ApxrguhNCD
   5H9ff9xc30+SJO2gGOGpp5KV6DvugAMPZNEhR1J0+L/z+svrWPLAUjZuHED1uLuaq9LpsDyFPn3e
   4Ygj9nL1WW1aY44efz595HiNa4tjjCObragQhgJ/qfqyM3BrjPEnNZ53hVqSpGx4662kLzqVYv2G
   jdzdbR9m5gzgHys/oaxsEJs3/47qI8CLaHhVOpkRPXWq/c9q3TJaoQ4h/AdwMZAfQlhc46lewONN
   U2L9YoxvAYc25/eQJEnbV1y8gD/84j5GvvpPxnzwCvtXruMvuYN5oM8+zP14CBtLx5ME5UD1anQ6
   XmxvVdowrfZhW2PzZgIPAj8lae9IJ/N1McbVzV2YJElqecXFC5g2bS7vLXuP/Zc8x3lla7gpvs88
   BvFTvsuDrGVz2U9hVXoV+gqSVeeiGu/S0OSO46pG3H3NIK12ZVs91J8AnwDnhxA6AZ+pur9HCKFH
   jPGdFqpRkiS1gOIHHuEP3/oTpyz/hPOYw+v0JcUVfJ3X+ZhfkITnn1bd3bnO58ZN7rjyyvMM02p3
   tnv0eAhhIvAj4H2SY8DTmq2HWpIktaClS3ltytXsf+vtXFu5JykGcDSLWEIK+CbVq8/1nV5Y32p0
   fUeAf90grXZru4Ea+E/gANs8JElqRz75BO66i9W//g05r7zCfPL5U+U5PMmNJIO18tk6NNe3Cj2a
   hlejndyhjqExgfodYG1zFyJJkprZ5s0wZw6kUmx+oJhHc3fnhk+HcU/FaDbxU5KWjsDWq86j2To8
   p1ehr2PPPTuzfv359O/fz9VodUiNGZv3R2B/oBjYVHU5xhh/2cy1basmx+ZJktQYMcKzzybzom+7
   jTV99+C363tz3Qf7s6JsILXH26VH242mesTdAmAeubkLycvLpW/fPNavr6gRnj2ERe1bRmPzanin
   6qNL1Udg66PIJUlSa/LOO3DrrcnphZs28fqRx/H93Y9n7pLdKC3djeogDdUr0ulgPA/4kM6dT2XY
   sPyq4PzfBmepAdsN1DHGohaoQ5IkZWrtWrj77iREL1oE55wDN95I8ZpyLvnPuZSUpGdFF1W9oKEN
   hcdVHboy3hAtNUJjpnzsCVwKDAe6VV2OMcYTm7MwSZLUCOXlMHdu0tLx4INQUADf+Q6MHUvx3/7J
   tCsf5PHHX2LDhr+w7SANjreTdk5jWj5uBW4HTiGZnfOvwAfNWJMkSdqWGOH55yGVovRPM3gr5nF3
   t6HcWP45Vv+tM5XzZhDjDZSXD2Hz5gtJOjehcUHaDYXSjmrMpsTnYoyfDSEsijEeXHXtmRjj51qk
   wvprclOiJKnjWbYs6YtOpWDDBt446jj+4/Fu/O3dccDNQD+qNxSm2zuuqHrx1VRvOqzebOiKtLRt
   TbUpMT3ZY1UI4RRgBdAn0+IkSVIjrFuX9EWnUmz659M81Duf23L34Z4PNrPhjtVUVj5AEpr7U/9R
   4J2BE6m7Ih1CCYcd1scVaakJNCZQXxNC2BX4L2A6sAvw3WatSpKkjqy8HB56KFmJLi6G447jjr5D
   +Wb5UD5ePoHq1ejcqhfU/M953aPAy6me3jEF6ATAoYd259lnr2vWH0PqKBoz5eP+qocfAwXNWo0k
   SR1VjLBwYdIXfdOfWFrRlbu6D+XG8sNZNW8TpaUfAvew9Wo01D7BsKFDWaoPY8nPv5yrrprQ/D+T
   1EE0ZsrHTXUuRYAY41ebpSJJkjqS5cth5kzWXvdbNr63mju6DOC36wp4JV4Ca9Mr0YHq/2TX/E93
   zdB8M/UfBT6P3NzF5OWdxd57D66aKT3GNg+pCTWm5aOY6oNcugFnkPRRS5KknbF+Pfz5z0lLx7PP
   8tyQA/nhyhE8tOl7xNLfkgzXqrkSXUTtFo60mgexdCYn5wm6d3+NGDeTk1MzQHsoi9SctjvlY6sX
   hJADPB5jPLp5SmpUDU75kCS1LRUV8Le/JSH6/vt5b//h/OqDPG58vwurP82jup2jM9VHgVP1+QqS
   1ej0seDpletrtrx9v37f5cYbzzA4S02sqaZ81LU/sMfOlSRJUgezaFESomfOhP79YcIEft5vBFdM
   e5ZNm/Zj63aO+laia4bpeVX3PUFe3hl067YLQ4f2dOydlEWNmUO9nuqWjwi8B/wwxnh3M9e2rZpc
   oZYktV4rVlT1Rf8fpas+4N6e+3DDp714qbI7mzaVU1nZExhB7XaO9CbDhlaik7nRublvMXJkbwO0
   1EKaZIU6xtiz6UqSJKl9mn33HB77r59w4vLXOKx8Nffm9GVG5WHM50ZiaarqrrobDCEJ0zU3F25v
   JfobBmmpldlmoA4hdAXGAQdV3bsYuCPGuK4FapMkqXWrqODJn05lxU+ncsL6lVSyF7/jG9zH22ys
   LKX25kLYeoMhbN3O8SE5OT9n8OB+DBvWj4kTxxugpVauwUAdQhgO3Af8A3iG5H+pjwcuqzox8Usx
   xl+1SJWSJLUCxcULmDJlBp1ffZ2zN77DOFaRSzcWMIT/4Aze59ckAXow9Y+5g+oV6fSYu5rj7dLt
   HN83REttyLZWqH8D/EeMcV7NiyGEk4CHSVarDdSSpHYrHaBff30lu3y6lvPjx/yBFexOGbdyMKMZ
   zcvswfZPKqwpvSJ9ETADuIAQytlnn+5MnWo7h9QWbStQD6wbpgFijA+FEMqAM5uvLEmSsqe4eAGT
   Jk1l5ZLNnM5aruFdjmIF97I//80Y5rM3lfwP1aPttndS4c1V1+quSJcycmRfNxhKbdy2AnUIIeTF
   GEvrXMwDNscYNzRvaZIkNb+aq9BlZZ0JcS3HVpRzBcs4nVU8xQBmcBVn8yKf0qXOq2uG55onFd7M
   1psL3yA39x1i/Apdu/bggAP6uMFQaicaHJsXQrgCOAr4Tozx7aprQ4FpwD9jjFe1VJH11ObYPElS
   RtKr0EuWRKAPw+lEIRsYz1/4kF1IcQizGMEqelF9uEpNV5OMskuPthtN0sKxAVhDly7ldOmyK7m5
   PZwTLbVhjRmbt8051CGE7wCXAj2qLm0Afh5jnN5kVe4EA7UkKRPFxQv4+tdvpnLVh1zARgpZTD9y
   uJW9SHEoL/IZardx1AzPaTXnQ88A3qdHj55VK8+GZ6m9yDhQ13ijXQBijGubqLaMGKglSTtt40Z+
   OPw0vvD2WxzDMu7jAGZwCH/nJiq5iq3bOGDr8PwWADk5PejWzRAttWdNFqhbGwO1JGmHVFbCggW8
   +z8/Z5e/zeOpyr7M4F+4h33ZsKUvuuZJhXXbOFYCOeTldWf48N0Nz1IHYqCWJHVI6Y2G4dU3OHvj
   Usaxio/pTIqhzGQoKzm46s66q9A1TypM+qGrR9p9zRAtdUAGaklSh1JcvIAff/tnHLV0GYWsYCAb
   mEk+KY5kEf2q7uoMnEhDq9C5uRXk5XVj770HM3BgLyZOPNkgLXVgTRKoQwg9gO8Be8UY/z2EsB9w
   QIzxgaYrdccYqCVJtWzcyJ0TvkP3ux/kmPghD3AAKUbwN4ZSQQ7VR35D0iOd3mSYnsrRhV69PmHW
   rO8ZniXV0phAva051Gk3Ac8C/1L19QrgLiBrgVqSJCor4bHHIJVi4y0z6VPagxkcxfkcxHq61vOC
   ugeuXAMk4blbt28ya5YtHZJ2TmMCdX6M8dwQwvkAMcYNIWwzpEuS1Hxeew1SKT79/R9YvraMmzb3
   YUbFsSzn6Kobytn6uG+o7pdO90hPATrRpcuLXHrpiYZpSTutMYG6LITQLf1FCCEfKGu+kiRJquPD
   D+G221gz/XrKl7zNzLgbN1UcyEKGkLRsHEj94+7Sah75PQP43xozoycZpiVlpDE91Om/GxtOcn7q
   McC/xhj/3vzlNViTPdSS1N6VlsIDD0AqxeaH/saDoQ/Xb9iLh9ifCj4F9qu6sTNJmG5o3N1GAHJy
   ejozWtIOa8qDXXYnOYYc4MkY44dNUN9OM1BLUjsVIzz+OKRSbJp1Gy/EHvy+tDe3l+/LOrpTO0Sn
   pcN07XF3sJn8/B6Ou5OUkYwCdQjhcKDmk+k3igAxxueaosidYaCWpHbmjTeSvujf3cjKTzZy0+a+
   3FyxH8sYCKwnCdJ1Q3RazTA9D+hETs5Cpkw5maKii1vsR5DUPmUaqOdTO1DXEmM8IaPqMmCglqS2
   b+6s+3n20p9w4opXGFK5gVnsTooDeI6hJCvMdVej64bougeyJGG6S5cXueyyEw3TkpqEB7tIklqN
   4uIFXDmCLscNAAAgAElEQVT5JvJfXcQ5ZW9zImv5K4NJsS/zGEj5Vn3RafVtNvRYcEkto6kOdukG
   XAwcS7Ji/Sjw2xhjaVMVuqMM1JLUdhQ/8Ah/+uaP+OKKpZzDShbRhxSDuJvjWUtejTu31dJR/6mG
   hmhJza2pAvWdwFrgFpI+6nFA7xjjOU1V6I4yUEtSG1BSwvyvTWLwI4+wiRxSHM6tjOAdelP/1Nbt
   tXSkNxuWk5/f3c2GklpEUwXql2OMw7d3rSUZqCWplfroI7jjDj6adh3xtTe4tXIAMyjgWQZTvbcd
   tn3wCrgaLam1aKqjx58LIRwdY3yi6k2PIjmKXJIk2LSJZ678OZ/85kY+98kyZtObFPsxh1Mp56Cq
   m+oG6PoOX5kDXEQSol8AFlWF6CGGaEmtWmNWqF8F9gfeJemh3gt4jeS3Y4wxHtzcRdZTkyvUkpRN
   MXLj179PnHEnZ5Sv4CV6k2IYdzGET9hM9Zi7hk4v9PAVSW1DU7V8DNnW8zHGt3e0sEwZqCUpS5Ys
   4bX/dzW5t93O5oocUuzPLfRnKYfWuKnmmLuGTy8MoQfdu/cyPEtq1ZrypMQ+wGBqtIg058EuIYQx
   wK+BTsCNMcaf1XneQC1JLWXNGrjzTj6a+hviq68xs3IQKfrzNMeT9EXX7R6suSrt6YWS2ramWqG+
   CvhXYAlQmb7eXAe7hBA6kbSUnAQsB54GLogxvlLjHgO1JDWnTZt4+upf8Mn03/H5j5cxl96k2JfZ
   DGIzw4BlwJCqmxvqj/bAFUltX1NtSjwPyI8xbmqasrbrCODNdCtJCOE24CvAK9t6kSQpQzHy+K9u
   YNlPpnHih29Syi7cwQGcwzF8XKsvGpKe54b6o2tuLvwt1WPuJrkqLaldakygfgnoA7zXzLWkDSTZ
   AJm2DDiyhb63JHUoxcULuP7S6/j8a09zQcVK9qQTc9ifyziZtzisxp11j/8+HpgP9KJ2gF4JvEUI
   i6r6o/vaHy2p3WtMoP4f4PkQwotAWdW1GGM8rZlqalQvR1FR0ZbHBQUFFBQUNFM5ktQ+FBcvYMqU
   Gbz++kq6fLqeM+NyJvAxN7OO2xnMRRzDUxzJ9vuibyb5VV0AFANLgYXk5nZj5EhH3Elq2+bPn8/8
   +fN36DWN6aF+heTv7F6kuoc6xhgf2Ykat19QMue6KMY4purry4DKmhsT7aGWpMZJh+iXX15CRdku
   jGEDhbzJaJbzEHuS4mj+Sqzqi27o6G+of1qHpxZKav+aalPi0zHGzzdpZdv+fp1JNiV+EVgB/BM3
   JUrSDhs37gfMmvU6n6OCQl7hfFbwBn1JMYI7GM4aelXd+TbJBsOGjv4Gg7SkjqqpNiU+GkL4CXAf
   1S0fzTY2L8ZYHkL4DklTXifgDzXDtCSptpqtHBs3llFZWclerGY8m3mZ98mlkhQHczRfYgl9a7wy
   HaDTGwxrhmj7oiWpsRqzQj2fevqam2tsXmO4Qi2pI6sZoDds+ATYA+hLLzZxNm8zgTcZyWru4GBS
   HMITDAIq6nmndIBeQ7K+sh+1D18J5OX1YPjw3Q3QkjqsJjvYpbUxUEvqaGr2QpeV7UoyfAk6sZZR
   dGICi/kSJTzM3qQYSTHd2MTeNd6hbgsH1G7jWEKydtKTnJyuDB3q4SuSBE17UuIpwHAgL30txnhl
   xhXuJAO1pI6kqOh6rr76YSoqegPrgX35LKsoZDHn8zJv0YcZHMwdHMRHdK961SvAt6kO0Vsf/Q2Q
   k9OTbt16evy3JDWgqTYl3gB0A04Efg+cAzwVY/xaUxW6owzUkjqC4uIFTJo0lSVLSoHDGMRaxvMy
   E1hIHuWkOJhbOIg3a/VFpw0AFgHjqNnCkYy229PwLEmN1FSBenGMcWQIYVGM8eAQQk9gdozx2KYs
   dkcYqCW1R/X1RvcicBbvUcgKDuE97mIYKUbyOINI5kXX18oByYr0QGABkEcIH3L++Qcxc+bP6rlX
   ktSQpprysbHq86chhIHAapLf1JKkDDS0ubATezKGdRTyBl/mTR5hN67nKB5gP8r4MkmAXlX1LnWn
   caR/Zb9NTk6PqnaOHlx55VddkZakZtKYQH1/CKEP8HPgOZJdK79v1qokqR2rbuWIJJsL+wE9OJRd
   KGQxF/AS77ArMziYSYxiNWtJfvV+wLbH2XlKoSRlww5N+QghdAXyYoyfNF9JjarDlg9JbVJR0fX8
   5Cfz2bQpAPsxkLWM50UKeZGebKrqiz6Y19m16hU1WzoCSYjuDHQmP7+nkzgkqZll1EMdQjgCeDfG
   uLLq64uAs0iO1CqKMX7UtOU2noFaUluSbu1YvPgVyst3pyf7cyZvUMgiPstK7uZAUozgMQYTSf/O
   9qhvSWoNMg3UzwNfjDF+FEI4Drgd+A5wGDAsxnh2UxfcWAZqSa3V1qcWbgT2JAc4ifcoZD2n8DqP
   MpgZHMz97EcZnfGob0lqnTIN1AtjjIdUPb4O+CDGWFT3uWwwUEtqbbbui04czDIK+YRxLGQ5fZnB
   CG5jOB9yBg0H6GRzYQg9qnqjnREtSdmS6ZSPTiGE3BjjZuAk4BuNfJ0kdRi1g3QusB/9Wcc4XmIC
   i+lNGbcwkhMZw2scRPVKtJsLJam92FYwngU8EkL4EPgUeBQghLAf8HEL1CZJrdLWx4Dn0oO9OYM3
   KGQWn2MFf2EYk/gSCxhY1Rf9BkmYrhmkfwtsJj9/d1s5JKkN2+aUjxDC0SR/Dzk3xrih6tr+QM8Y
   43MtU2K9ddnyIanFNDQvOoe1nEguhbzEabzG4wxiBgdzH/tRSm7Vq9Mr0muAPUlOLpwHdAIWcsEF
   +3vYiiS1Yk1yUmJrZKCW1Ny2XoVO90WvZwS9mcBixvESq+i5pS/6fc5k61MLa/ZGLyE5tbAHPXtW
   8r3vfZGiootb7oeSJO0wA7Uk7aCtNxeuB/ajX1VfdCEvshsbuYWRpDiIV9iD+kfcVZ9amJPTs+rE
   QjcXSlJbY6CWpEaoryca9qM7mzidNylkEUeynHsYxgwO4hH2ruqLdla0JLV3BmpJqiMdnt9+ez1l
   ZRspL1/Dpk27k16NziGfApZRyCJO51WeYCAzOJh72Z+N5FJ7XrRBWpLaOwO1JNVQfez3fiRB+GbS
   LR3D+YBCXuJCFvE+3UkxklkcxHtb9UXXbevIIS+vO8OH7247hyS1QwZqSR1W/ScW7g6MBK4GruAz
   rOcCXqGQRezJBm5lBClG8BIXUv8qtAeuSFJHY6CW1OE0dGJhshJ9IN3YzFcYQSH/j3/hXe5lf1KM
   5O/sTSVjqH9zoavQktRRGagldQjFxQuYNm0ur776Gu+8U0F6U2FaIHI8yyjkBc7gdf7JccygG/ew
   P59yCvZES5IaYqCW1O4VFV3PtdcuYuPGccB1JEE6OQT2QD6gkIWMZzEfkUeKs5nJR6xiH6p7qF2N
   liQ1zEAtqd2qbu2oAO4BrgA6swcbuICFFLKYAazf0he9mEKSY79rrkCvoUuXcrp02ZXc3B4MHdrT
   EC1JqqUxgbpzSxUjSTur4VF3ucBI8tjIabxKIQs5lne5n+O4jEN4mP5UkgNUUh2m5wGD6NLlRS67
   7DRPKpQkZcwVakmtVu0NhgdSc9RdYF++wHIKeYEzeYdn2JUUF/AX3mUDg/HEQklSU7DlQ1Kb0/AG
   w2TU3QF8SCEvMZ7FrKMLMziLmeSygrOp29IRQjn77OOmQknSzjNQS2oTah/9PQT4V2puMNydDZzP
   XhTyUwazlpkcRIoRLKQQmEt1K8f7wDL23nsAw4b1Y+LEkw3SkqSMGKgltTpbH7hSCexK+uhvuB24
   gq7AqbxGIQs5jnd4gLNJUcHfGErFVvOi5wGdqvqiT7QvWpLUZAzUklqNdJB+8cV1bN7cs+pqP6D6
   d1SgE8dwEhO4hLN4hefoR4oL+DPLWL+lL7ruqDvnRUuSmo+BWlJW1DeVY/PmfYlxQJ07rwaK2I/V
   FLKQC1nEBgaRYk9u5WKW8yBbh2dH3UmSWo6BWlKL2XozYe2pHEkrR9GW+3fjU85jbwr5H4bwCbM4
   iBn8Oy+wBBiDGwwlSa2Bc6glNauaIXrZsl2orLwIeAM4gPRUDuhP+ldNF0o5hTcpZDEnsJRizuTH
   XMI8XqGCHOAtkjA9D/iQEK5lr736M2zYvm4wlCS1WgZqSTul9pHfbwB/IAnQB9a4qzMQOYa3KOSb
   nM1tLKQHKfKZwBGsYy/gUmAByUr0k+TkLK4xK/pSQ7QkqdUzUEtqtJoTOjZsyKX6yO90iO4MlAOQ
   z5sU8jcu5EXK6E2KoRzGYt7lbZLwvARYQU7OaTUC9I8M0JKkNsdALale6XaO5cs/YOnSdykvr6S0
   tAcwjGSjYPrXR3WI7sM6ziNSyK3k8ytm8WXOpS/PcRBJK8fvScbbfcxll53teDtJUrtgoJa0leLi
   BVxyyRxKStKbCg+jerxdMpkjHaK7UMqX6ckEDuBElvMge3EN32Mur1HORmAdXbr8ky5dXq8xlWOS
   K9GSpHbDQC1pK1Om3EZJyfVUbypMh+i0zRzNQAr5LOewhBfpTYrz+VcqWctacnIeY/Dgfm4mlCR1
   CAZqSVuke6RfeOHTqis1f0WUsw9ruJAfU8jNlLORFGfyOY5hKZvIyXmZwYP7cfSwfkycON4QLUnq
   MAzUUgdS98CVTp060bdvN1avXkN5eSVlZYOIsV+NV5SzKxs5lxso5D725w1uYwQX8Bee4VPgIXJz
   3+KzI3tz5ZXfN0RLkjokD3aR2rG6Gws3buxBRcX+JAempA9OSR/lHUi3duTyBb7MbynkA07iCeaQ
   T4prmU0PypkJvE+PHunJHJ5SKElqvzwpUerAam8snEPtTYVX1PicvvYjjuRLFPIdzmUprzCAFIO4
   kz34hA/Iycmhd+++HvUtSepQPClR6gAaauNYvnwt5eX3k4Tma6i9qbB65N0Q1nAhV1HIb4BZzOA4
   juBI3ua6LXfn51/O1KljDNGSJNXDQC21Uekg/eKL69i8eV9gAuk2jnXr5gC5VXem/zUv3/La3qzn
   HH7PBP7IMFZzO/9GIdfwT94B/ofk5MIp5OUtZfjwXq5IS5K0DQZqqQ2qbufoR9L/nG7fuKbOZ0gH
   6c6cyBh+wgQOZBTvMo9B/C/f4kFeZzO9gW+RDtK5uW8xcmRvrrzy6wZpSZK2w0AttQE1NxeuWvUx
   ZWUVrF9/Fw21cSRGAZfzefamkM9xHu/wOv1IsRvfZAifdu1Ely7PMqBvHqtXvwCcXePglW8YpCVJ
   aiQDtdTKbb258Aaqg3R5jTvLt3zei6VcyKMUciudWccM9uHYTiPoeuDBDBzYi5SHrUiS1GRaXaAO
   IRQBXwc+qLp0WYxxdvYqkrJr2rS5lJTUbOWA6vA8imTs3WR24RjOYSyFLOcgfskd/Cv/xm08yVHk
   5092U6EkSc2k1QVqIAK/jDH+MtuFSC2lbktHz56dWb++nJ49O7N8+aaqu2r+6zoKmExnihjNPyjk
   d4zhWv6esxs3dBnCM3uexkcbPqJ//9sZPfCvTJxomJYkqbm0xkAN1QNzpXarZohesiSwceM4kpaO
   8axenUzrSD6n/3VIr0pHDqc7hbzG+fRhaafuPHnA53hqyjROP/8UTs/GDyNJUgfW6g52CSH8CPg3
   4BPgGeC/Yowf17nHg13UpqTDc1lZZ9auXcYnn6xjxYq+lJb+ltqHrNT3eQEwh8EcyniuZwKr6MIm
   UhTy6F6r+N7141x9liSpmbTag11CCPNIZn3VNRn4LXBl1ddXAb8AvtZCpUlNrnpT4TWkw3H1qDvY
   ejpH9ederOUsSpjAXzmYX3EnQ/hWp758NOxzDBxUwfcmGqYlScq2rATqGOPJjbkvhHAjcH99zxUV
   FW15XFBQQEFBQVOUJjWp4uIFXHTRdaxefXvVlblsfWphea3PndjEKP5KIXfyJX7NfAr4DVfwAKew
   ia6MPmkKi2df1WI/gyRJHcn8+fOZP3/+Dr2mNbZ89I8xrqx6/F3g8zHGcXXuseVDrdbWvdH9qA7Q
   RVUf6XYOSFatZ3MY+1DI77iAEt4mjxTncjuwml9teW+PAJckqWW12paP7fhZCOFQkmkfbwHfzHI9
   UqPVbu+o2QudVnPc3WQG8h+M5wkmMIse4QPu3WUoXwlH8V7vXVm/fiU9e3aC9efTv38/Bg7s5bQO
   SZJaoVa3Qt0YrlCrtRo9+grmzk2vPBdVfaT7ppMe6p7cx1mMoJDpHMor3NNpIP/IH8GZ/3sJY08t
   yErdkiSpfm11hVpqs8rKav4rlV6NPo5OlHMS4yjkRcbyCgt7D+LhgZ/n1wNG8a3//BJ/cNVZkqQ2
   y0At7aS6o/CgC2+//VGNO07mEL5OIb0Zx0zeZTAP7t6Pp6dezcnjTuP4bBUuSZKalC0f0g7YesPh
   /1G3pWMAdzGeQRSSohcfcnvnPjw+9HNs2mdvJk482R5oSZLakMa0fBiope2oP0TXnNJxBT34IWfy
   ZwpJcThP8WfyuafnnlT8y+f5zqRRhmhJktooe6ilBtRs1+jatZyjjx7AE0+sqNW+scsue7J27TJW
   rtyFVat+Se0Q3ZkcKvgif6OQ+zmV3/AYx3IjX+c07qOUbhx/eBHz5xRl7WeUJEktw0CtDqFuv3N1
   SAZYwMMPz6S8vG77BtQN0QAjWUQhcxjP71jOQFL047+YxwfsWet75uVVNPvPJUmSss9ArXZp2wG6
   ZkgGmFsVppPH1WEa0v+K9GcF43iMQg6hD2u4heP4IgfzKjeQhPCptV6Xn385EyeOaaafTpIktSYG
   arULOxag6/5j37nex93ZwBk8TyGjOYJ/8heO5j8ZwSOkiOQAC+jW7Tzy8/uTm7ueEL5Nr157kJdX
   4QEskiR1IAZqtTn19T/fcsvyqtMJYfsBurzBr3PYxAk8RCEpvsK9/INh3EQ/zmA5G+lOEqIvID+/
   f9XJhd82OEuS1MEZqNWm1D7aO/Hoo+excePtNe7aXoBOjv2ubtEYxSGdzuWCiqGM5ybe4yZS/IBL
   uZb3+Qz9+n2VAwd8v8bqsyFakiRVc2ye2pTaR3unFVV9pNVdoa670RD69fsqI/eIfGXDCkZ/sJA9
   QhmzdxvOg7sdwqKKTYTQtUaAdna0JEkdlWPz1ObVbe9YsWJ9PXdtbwX6OPr1+xMDBnyb3bvvygmf
   vMxXu6xgz5LX4StfgcJLoaCA8zp14rxm/WkkSVJ75Aq1WoW6wXnSpFEAW7V3dOtWt70Dkr7mmVUH
   riT69fsqAwZ0o1evPejWtZwfHb87R72xEO65B446CgoLkzDdo0dL/HiSJKmN8qREtTqNDc75+ZPZ
   ZZc1PP/89XXeYevwnJ9/ORdeOIgnn1xJaWmn6jaNIbtBKgW33gq7756E6AsugP79W+JHlSRJ7YCB
   Wq1KfRsKGw7O0KfPRaxZc/NW1w866JsMGrRn7fCc7nF+7z2YNSsJ0qtWwfjxSZAeObLZfi5JktR+
   2UOtVmXatLm1wjRASck19OlzUQOvKKv36qBBezJ79lXVFzZuhNtuS0L044/DaafBz34GJ5wAnTo1
   UfWSJEn1M1CrxZSVNfSPW/3BeciQnvTtO7nOinbVCYSVlbBgQRKi//xnOOKIZCX6jjvsi5YkSS3K
   QK0mVV+PdLodo2vXutM4Eg0F56uumgDA9OlTtrR3XHb6MI5/fDZcXAi77pqE6JdeggEDmv+HkyRJ
   qoc91GoyDfVIT506mrFjj2vg+cuZOnUMANOnz6u/L/r996tbOpYvh3HjkiB9yCEt+vNJkqSOx02J
   alH1H7oCo0dP2dLzXFy8oOHgXNPGjXD//UmIfvRROPXUJER/8Yv2RUuSpBbjpkS1qIZ6pEtLqwPw
   2LHHNXzqYGUlPPYYzJiR9EUffngSomfNgp49m6NkSZKkjBmo1WQa6pHOy6vY9gtfey1Zib7lFujV
   KwnRixbBoEHNUKUkSVLTMlBrh2xr0+GkSaMoKWlgKkddH35Y3Rf9zjvJgSv33JP0RYdt/q2KJElS
   q2KgVqPVt6mwpGQyULuVo+ZUjokTx1S3eJSWwgMPJC0djzwCp5wCP/4xnHQSdPYfRUmS1Da5KVGN
   1phNh1uJMemLTqXg7rvh0EOTlo4zz4RddmnmiiVJkjLjpkQ1qcZsOtzijTeq+6K7dUtC9AsvwODB
   zVylJElSyzJQq9G2u+lw9Wq4/fYkSC9ZkvRF33UXHHaYfdGSJKndysl2AWo7Jk0aRX7+5FrXDtzn
   Uq45vCucfjrss08yM3rKFFi2DH79a/jsZw3TkiSpXbOHWsC2p3fUvW/6tLnkv7+SL3/4Aid/8iZd
   Dv9s0tJx1lnQu3cWqpckSWoe9lCrUbY3vaPGRcY+/TBjS26D3Fz4j0IYPx723rulS5YkSWo1XKHW
   tqd3zPwu3HFHMuruzTeTvujCwuQUQ1s5JElSO+cKtRql7vSOXDbxZf7K9569C4ZOgzFj4PLLYfTo
   ZGVakiRJWxioVTW9I3IUT1JIinO5g5c4iOc+k89xjz0Bu+6a7RIlSZJaLQN1O9XYTYYsWcK0PVbQ
   NXc3yjbvwQwm8DmeoXP+75j6szGGaUmSpO2wh7odqm+TYX7+ZKZOHZ2E6jVr4M47k77o116D88/n
   saEjuHrOu5SWda46Mvzk+gO4JElSB9KYHmoDdTtU3ybDXDZx+WGFFO1TAfPmwahRyebCMWOgS5cs
   VSpJktS6uSmxg6reZBg5gn9SSIrzuJ2VJd3hW5Ph97+HPn2yWqMkSVJ74UmJ7dBelR8ymat5lWHc
   woW8z54cyVNcevQE+MY3DNOSJElNyBXqVqrRmwrTPv4Y7roLZszg9wsXcecu+Vy09mae4kggkJ9/
   ORMnjmmx+iVJkjoKA3Ur1OiTCzdvhtmzIZWCOXPgpJPge9+j65e/TJ95T7Lr9GKOL51dtclwjJsM
   JUmSmoGbEluhbZ5c+OCV8MwzSYi+7TbYb79kc+G550LfvlmoVpIkqf1yU2IbVffkQoC9WMpZr/0D
   DjwQysuTEP3EE5Cfn4UKJUmSlGag3gE73Ne8k5KTC6EXazmbu5jADEbwIk9W7A1//CMcfTSEbf6P
   kiRJklqIgbqRGt3XnKnNm7n66F5M/MdwvrB+OQ9zItOYxKv7PMnPp50C//IvTfe9JEmSlDF7qBtp
   m33Ns6/K7M1jhOeeS/qiZ82CffZh8WeP5spXAh9U9vLkQkmSpCyxh7oJ1dfXDFBa2mnn3/Tdd+HW
   W5MjwEtLk77oxx6D/fZjJHDnzr+zJEmSWkhWDnYJIZwTQngphFARQvhsnecuCyG8EUJ4NYQwKhv1
   1Sfd11xXXl7Fjr3R2rVw001w4olw6KHw1lvwu99BSQn8+MfJ1A5JkiS1Gdk6KXExcAawoObFEMJw
   4DxgODAGuD6E0CpOc5w0aRT5+ZNrXUsOSzl5+y8uL4cHH4Rx42CvveCee+Dii2H5crjhBjj2WDcZ
   SpIktVFZafmIMb4KSU9KHV8BZsUYNwNvhxDeBI4AnmzZCreW7l+ePn0KpaWdtn9YSozwwgtJX/TM
   mbD33jBhAkybBrvv3oKVS5IkqTm1th7qAdQOz8uAgVmqZStjxx63/Y2By5YlfdGpFGzYABdeCI88
   Agcc0DJFSpIkqUU1W6AOIcwD+tXz1OUxxvt34K1a/xiSdevgz39OQvRzz8FZZ8H11yetHDmtomNF
   kiRJzaTZAnWMsRHNxVtZDgyu8fWgqmtbKSoq2vK4oKCAgoKCnfh2GaiogIceSkL0Aw/AF74A3/wm
   nHoq5OW1bC2SJElqEvPnz2f+/Pk79JqszqEOIfwd+O8Y47NVXw8HZpL0TQ8EHgL2rTt0OhtzqLd4
   9VW48cakL3rQoGTU3fnnwx57ZKceSZIkNZtWO4c6hHAGMA3YHSgOITwfY/xSjPHlEMIdwMtAOXBx
   9pJzA557Drp2hYcfhmHDsl2NJEmSssyTEiVJkqQGNGaF2h1zkiRJUgYM1JIkSfr/7d15jF1lHcbx
   7wNYUQEVicGlpmIqCCqLYABxiVjjEi1EjVvcNcYFiAJxqXGJG0rcDUET3Bci7opIEGkqFlFri5QC
   4kJUFNwK0kAV259/3HdgOt6ZucOhc+bW7ye5mbPe83vvyUyfvvc956gDA7UkSZLUgYFakiRJ6sBA
   LUmSJHVgoJYkSZI6MFBLkiRJHRioJUmSpA4M1JIkSVIHBmpJkiSpAwO1JEmS1IGBWpIkSerAQC1J
   kiR1YKCWJEmSOjBQS5IkSR0YqCVJkqQODNSSJElSBwZqSZIkqQMDtSRJktSBgVqSJEnqwEAtSZIk
   dWCgliRJkjowUEuSJEkdGKglSZKkDgzUkiRJUgcGakmSJKkDA7UkSZLUgYFakiRJ6sBALUmSJHVg
   oJYkSZI6MFBLkiRJHRioJUmSpA4M1JIkSVIHBmpJkiSpAwO1JEmS1IGBWpIkSerAQC1JkiR1YKCW
   JEmSOjBQS5IkSR0YqCVJkqQODNSSJElSBwZqSZIkqQMDtSRJktSBgVqSJEnqwEAtSZIkdWCgliRJ
   kjroJVAneVaSy5JsSXLIpOVLktycZG17ndZHfZIkSdKo+uqhvhQ4Flg1ZN2vq+rg9nr1PNelebBy
   5cq+S1AHnr/x5bkbb56/8eW52/H1Eqir6oqq+lUfx1b//MMy3jx/48tzN948f+PLc7fjW4hjqB/Y
   hnusTHJU38VIkiRJM9lle71xkvOAvYesenNVfWea3f4ELK6qjW1s9TeTHFBVN26vOiVJkqQuUlX9
   HTy5ADixqn4xl/VJ+itakiRJ/1eqKjOt32491HNwa4FJ9gI2VtWWJPsAS4HfTt1htkZJkiRJ86Wv
   2+Ydm+QPwOHA2UnOaaseC1ySZC1wFvDKqrq+jxolSZKkUfQ65EOSJEkadwvxLh8jSfLOJJckWZfk
   /CSL+65Jo0tyapLL2zn8epK7912TRjPdg5m0sCV5UpIrklyV5A1916PRJflUkuuSXNp3LZqbJIuT
   XHC6k3MAAAb6SURBVND+Zq5PcnzfNWl0SXZNcnHLmhuSvHfabce1hzrJ7hN3/0hyHHBgVb2857I0
   oiTLgPOramuSUwCq6o09l6URJNkP2Ap8ghkuKtbCkWRn4ErgCcA1wM+A51bV5b0WppEkeTSwCfhc
   VT2s73o0uiR7A3tX1bokuwFrgGP83RsfSe5aVTcl2QW4EDipqi6cut3Y9lBPuZXebsDf+qpFc1dV
   51XV1jZ7MXD/PuvR6Hww01h6JIOn0F5dVbcAZwLLe65JI6qqHwEb+65Dc1dV11bVuja9CbgcuG+/
   VWkuquqmNrkI2Bn4x7DtxjZQAyR5d5LfAy8CTum7Ht1uLwW+13cR0g7sfsAfJs3/sS2TNE+SLAEO
   ZtCJpDGRZKck64DrgAuqasOw7RbCbfOmNdvDYapqBbAiyRuBDwEvmdcCNaNRHu6TZAXw76r60rwW
   pxndzgczaeEaz7F90g6iDff4KnBC66nWmGjfph/UrvU6N8njqmrl1O0WdKCuqmUjbvol7OFccGY7
   f0leDDwFOHpeCtLI5vC7p/FwDTD5wu3FDHqpJW1nSe4EfA34QlV9s+96dPtU1Q1JzgYOBVZOXT+2
   Qz6SLJ00uxxY21ctmrskTwJOBpZX1ea+69Ht5kOWxsPPgaVJliRZBDwb+HbPNUk7vCQBzgA2VNWH
   +65Hc5NkryT3aNN3AZYxTd4c57t8fBXYF9gC/AZ4VVX9pd+qNKokVzEY4D8xuP+iqnp1jyVpREmO
   BT4K7AXcAKytqif3W5Vmk+TJwIcZXFRzRlVNe/snLSxJvszgwWf3Av4CvLWqPt1vVRpFkqOAVcAv
   uW3o1Zuq6vv9VaVRJXkY8FkGHdA7AZ+vqlOHbjuugVqSJElaCMZ2yIckSZK0EBioJUmSpA4M1JIk
   SVIHBmpJkiSpAwO1JEmS1IGBWpIkSerAQC1JQyS5V5K17fXnJH9s0xuTXDbPtSxP8pBJ8+9IMucn
   jLYHu1w6zboDkvwwyRVJfpXkLV1qnqGGYW15fJtemeQR2+O4krQ9GaglaYiq+ntVHVxVBwOnAx9s
   0wcBW+/o4yXZeYbVxwL7T6rtbVV1/h147LsA3wLeU1X7AQcCRybZHg9bGtaWH07MctvDLyRpbBio
   JWk0mfRz5ySfTLI+yblJdgVI8qAk5yT5eZJVSfZty5e03t9LkvwgyeK2/DNJTk/yE+B9w/ZPciTw
   NODUJL9Isk/b7xntPQ5L8uMk65JcnGS3drxVSda01xGztO15wIVV9QOAqroZeC3whnaMtyc58dYP
   YtDuB7Tpb7R61yd5xaRtNiV5V6vroiT3nq0t23zYyROTrG71fyXJ3dryU5Jc1j7LoU8sk6T5ZqCW
   pLlbCny8qh4KXA9MBMJPAsdV1aHAycBpbfnHgE9X1YHAFxk8un3CfYEjquqkYftX1Wrg28BJVXVI
   Vf2W1pObZBFwJnB8VR0EHA3cDFwHLKuqRwDPmXK8YfYH1kxe0I6ze5Ld+N9e48nzL231HgYcn+Se
   bfldgYtaXauAV8zUlslvnmQvYAVwdGvDGuD1SfYEjqmqA9pn+c5Z2iVJ82KXvguQpDH0u6r6ZZte
   AyxpPahHAmclE53ZLGo/DweOadNfAN7fpgs4q6qqBdcjptkfbushnzy/L/DnqloDUFWbAFrQ/niS
   A4EtwINHaNPU959wp1n2OyHJRNsWM/jPxk+Bf1fV2W35GmDZCMeaWHc4g5C/un0Wi4DVwA3A5iRn
   AN9tL0nqnYFakubuX5OmtwC7MvjGb2MbZz3MdCHypvZzJ+D6GfYfNrZ4uvHGr2MQtF/QxmZvnma7
   CRuAx2xTbLIPcFNVbUzyH7b9RnNiiMvjGPSKH15Vm5NcMLEOuGXS9lvZ9t+bUcZJn1dVz5u6MMkj
   2zGfyWBYypwvzpSkO5pDPiSpu1TVjcDvkjwTIAMPb+tXMxh6AfB8BkMgtlFV/5xh/xuBPabuAlwJ
   3CfJoW2f3VuA3gO4tm33QmCmCx5hMAzlqIk7h7SLFD8CfKCtvxo4pK07BHhgW74Hg/9EbE6yH4Oe
   5dkMa8vUdv0EeFSSB7Vj3i3J0vYtwD2q6hzg9QwunpSk3hmoJWk0Nc305PnnAy9Lsg5YDzy9LT8O
   eEmSS9o2J0zzXtPtfyZwcrtAb59bd6y6BXg28LG2z7nAnRmM3X5RW7YvsGmG2qmqze1YK5JcCfwV
   uKqqPtQ2+RqwZ5L1wGsYBHmA7wO7JNkAvBe4aJrjTB4nPbQtU+r5G/Bi4MvtM1vd2rE78J227EcM
   euIlqXep8g5FkqTbJFkOvBV4alVdO9v2kvT/zkAtSZIkdeCQD0mSJKkDA7UkSZLUgYFakiRJ6sBA
   LUmSJHVgoJYkSZI6MFBLkiRJHRioJUmSpA7+C2ZmelBvYQMTAAAAAElFTkSuQmCC
   "
   >
   </div>
   
   </div>
   
   </div>
   </div>
   
   </div>

   <script src="https://c328740.ssl.cf1.rackcdn.com/mathjax/latest/MathJax.js?config=TeX-AMS_HTML"type="text/javascript"></script>
   <script type="text/javascript">
   init_mathjax = function() {
       if (window.MathJax) {
           // MathJax loaded
           MathJax.Hub.Config({
               tex2jax: {
               // I'm not sure about the \( and \[ below. It messes with the
               // prompt, and I think it's an issue with the template. -SS
                   inlineMath: [ ['$','$'], ["\\(","\\)"] ],
                   displayMath: [ ['$$','$$'], ["\\[","\\]"] ]
               },
               displayAlign: 'left', // Change this to 'center' to center equations.
               "HTML-CSS": {
                   styles: {'.MathJax_Display': {"margin": 0}}
               }
           });
           MathJax.Hub.Queue(["Typeset",MathJax.Hub]);
       }
   }
   init_mathjax();

   // since we have to load this in a ..raw:: directive we will add the css
   // after the fact
   function loadcssfile(filename){
       var fileref=document.createElement("link")
       fileref.setAttribute("rel", "stylesheet")
       fileref.setAttribute("type", "text/css")
       fileref.setAttribute("href", filename)

       document.getElementsByTagName("head")[0].appendChild(fileref)
   }
   // loadcssfile({{pathto("_static/nbviewer.pygments.css", 1) }})
   // loadcssfile({{pathto("_static/nbviewer.min.css", 1) }})
   loadcssfile("../../../_static/nbviewer.pygments.css")
   loadcssfile("../../../_static/ipython.min.css")
   </script>