How to return an array column in numpy?

Question

So I want to write a function that performs z-score transformation for each element in an array and the function should return an array of the result. When I perform the transformation to every element in the array using a for loop its not a problem.

   x1 = np.array([[4,3,12],[1,5,20],[1,2,3],[10,20,40],[7,2,44]])
   mean = np.average(x1)
   std = np.std(x1)

   for i in x1:
       arr1 = ((i-mean)/std)
       print(arr1)

   type(arr1)

However when I use a for loop for the column 1 of the array I get float values.

for x in x1[:,1]:
    arr2 = ((i-mean)/std)
    print(arr2)
    
type(arr2)

What can I do to make sure arr2 that is returned is a 1 dimensional array.

Please avoid posting pictures on SO and rather post the code/data itself here. Thank you. — Ehsan
– Ehsan, Commented Sep 1, 2020 at 0:12
did you mean to calculate the mean and average of columns/rows separately? the question is not clear to me. Maybe you can elaborate on it. Thank you — Ehsan
– Ehsan, Commented Sep 1, 2020 at 0:22
Thanks for the response Ehsan I have posted the code as well. No I just want to retrieve the "transformed" column/rows not calculate the mean of columns and rows separately. — LordVoldemort
– LordVoldemort, Commented Sep 1, 2020 at 0:26
Does the posted answer solves the issue? if not, could you please provide us with your desired output? — Ehsan
– Ehsan, Commented Sep 1, 2020 at 0:37

Ehsan · Accepted Answer · 2020-09-01 00:21:41Z

1

You do not need a loop to calculate that:

(x-x.mean())/x.std()

Example:

x = np.arange(3*4).reshape(3,4)
#[[ 0  1  2  3]
# [ 4  5  6  7]
# [ 8  9 10 11]]

(x-x.mean())/x.std()
#[[-1.59325501 -1.30357228 -1.01388955 -0.72420682]
# [-0.43452409 -0.14484136  0.14484136  0.43452409]
# [ 0.72420682  1.01388955  1.30357228  1.59325501]]

Then you can select any columns/rows you would like:

selecting first column:

((x-x.mean())/x.std())[:,0]
#[-1.59325501 -0.43452409  0.72420682]

answered Sep 1, 2020 at 0:21

Ehsan

12.5k2 gold badges24 silver badges36 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

LordVoldemort Over a year ago

Thanks Ehsan, I think your code gets me the output I want.

Alex Rancea · Accepted Answer · 2020-09-01 00:29:19Z

0

If you want to have the same structure you presented, this can help you

a = np.array([])
for i in x1[:,1]:
    val = ((i-mean)/std)
    a = np.append(a,val)
type(a)

answered Sep 1, 2020 at 0:29

Alex Rancea

973 bronze badges

2 Comments

mathfux Over a year ago

We are working with numpy, so looping should be avoided here as much as possible. Another answer illustrates a way to do it.

Alex Rancea Over a year ago

This is why I said ‘the same structure’. You’re right

Collectives™ on Stack Overflow

How to return an array column in numpy?

2 Answers 2

1 Comment

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related