CODEWITHSUNDEEP
pythonnumpy
cpc333
cpc333

Reputation: 1573

Numpy and += effect

Something I don't understand with Python 3.3.3 and NumPy:

from numpy import *
x1 = zeros(1)
x2 = x1
x1+=1

It turns out, this makes both the x1 and x2 variables [1], which I don't understand. If you instead do x1=x1+1, then I get x1 as [1] and x2 as [0], which is what I was after.

Upvotes: 3

Views: 94

Answers (1)

Konrad Rudolph
Konrad Rudolph

Reputation: 545548

x2 = x1 makes both variables x2 and x1 refer to the same object.

x1+=1 changes the object underlying the reference x1 (and x2).

Conversely, if you did x1=x1+1 then you create a new object (the result of x1+1) and assign the result to x1 only, while x2 remains unchanged and refers to the original object.

This is unrelated to NumPy, by the way – it’s a consequence of the general way references work.

Upvotes: 6

Related Questions