numpy.logaddexp2#

numpy.logaddexp2(x1,x2,/,out=None,*,where=True,casting='same_kind',order='K',dtype=None,subok=True[,signature])=<ufunc'logaddexp2'>#

Logarithm of the sum of exponentiations of the inputs in base-2.

Calculateslog2(2**x1+2**x2). This function is useful in machinelearning when the calculated probabilities of events may be so small asto exceed the range of normal floating point numbers. In such casesthe base-2 logarithm of the calculated probability can be used instead.This function allows adding probabilities stored in such a fashion.

Parameters:
x1, x2array_like

Input values.Ifx1.shape!=x2.shape, they must be broadcastable to a commonshape (which becomes the shape of the output).

outndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must havea shape that the inputs broadcast to. If not provided or None,a freshly-allocated array is returned. A tuple (possible only as akeyword argument) must have length equal to the number of outputs.

wherearray_like, optional

This condition is broadcast over the input. At locations where thecondition is True, theout array will be set to the ufunc result.Elsewhere, theout array will retain its original value.Note that if an uninitializedout array is created via the defaultout=None, locations within it where the condition is False willremain uninitialized.

**kwargs

For other keyword-only arguments, see theufunc docs.

Returns:
resultndarray

Base-2 logarithm of2**x1+2**x2.This is a scalar if bothx1 andx2 are scalars.

See also

logaddexp

Logarithm of the sum of exponentiations of the inputs.

Examples

>>>importnumpyasnp>>>prob1=np.log2(1e-50)>>>prob2=np.log2(2.5e-50)>>>prob12=np.logaddexp2(prob1,prob2)>>>prob1,prob2,prob12(-166.09640474436813, -164.77447664948076, -164.28904982231052)>>>2**prob123.4999999999999914e-50
On this page