Arithmetic mixin

Overview

The NDArithmeticMixin adds methods for performing basic operations on NDData objects: add(), subtract(), multiply() and divide().

The operations are permitted only if the two operands have the same WCS and shape and the units, if any, consistent with the operation performed. The result is masked at a particular grid point if either of the operands is masked at that point.

The operations include a framework to propagate uncertainties that are based on the classes NDUncertainty.

Warning

Uncertainty propagation is still experimental, and does not take into account correlated uncertainties.

Usage

As with other mixins, using the arithmetic mixin starts with defining your own class:

>>> from astropy.nddata import NDData, NDArithmeticMixin
>>> class MyNDDataArithmetic(NDArithmeticMixin, NDData): pass

Then, create instances of this new object with your data the way you would with a plain NDData object:

>>> ndd1 = MyNDDataArithmetic([1, 2])
>>> ndd2 = MyNDDataArithmetic([3, 4])

The mixin adds methods on these instances for combining them:

>>> ndd1.add(ndd2)
MyNDDataArithmetic([ 4.,  6.])
>>> ndd2.multiply(ndd1)
MyNDDataArithmetic([ 3.,  8.])

One important note: the order you list the mixins and NDData matters; the base class, NDData should be on the far right.