CODEWITHSUNDEEP
pythonnumpy
dineshdileep
dineshdileep

Reputation: 805

In Numpy, how do I find element wise inverse (reciprocal) of a given vector?

enter image description here

This shows my numpy session. I wish to find the reciprocals of the elements of a given vector and store it in another vector. The function power in numpy is not helping.

Update This worked for me. 

numpy.power(x*1.0,-1)

Code sample

import numpy
x = numpy.arange(5) + 1
print (x)
print (numpy.power(x, 2))
print (numpy.power(x, -1))

Upvotes: 1

Views: 4433

Answers (2)

Thomas Baruchel
Thomas Baruchel

Reputation: 7517

As stated by Maciej Lach, your input method has to contain floats (otherwise output will be cast to integers by the numpy.reciprocal method). But you can achieve it even with integers with:

1./x

(because 1. is a float which will promote your results to floats).

Upvotes: 3

Maciej Lach
Maciej Lach

Reputation: 1691

Numpy provides a function to calculate reciprocal of the vector:

import numpy

x = numpy.arange(5) + 1
print (x)

r = numpy.reciprocal(x.astype(float))
print (r)

Which gives output:

[ 1.  2.  3.  4.  5.]
[ 1.          0.5         0.33333333  0.25        0.2       ]

Note that your input method has to contain floats, otherwise output vector will be cast to integer and you will end up with: [1 0 0 0 0].

Upvotes: 4

Related Questions