Reduce dimension of numpy array by defined function

Question

With a self defined function:

    def myfunc(x1,x2, ... ,x10):                 # list
        ... 
        matrix operation on x1, x2 ..., x10      
        ...
        return X                                 # one value

What is the right way to use array operation to transfer A to B by calling myfunc() like this:

    A.myfunc() >> B

Joran Beasley · Accepted Answer · 2013-11-01 15:51:56Z

1

import copy
def myfunc(x1,x2, ... ,x10):                 # list
    ... 
    matrix operation on copy.deepcopy(x1), copy.deepcopy(x2) ..., copy.deepcopy(x10)      
    ...
    return X      

B = myfunc(a,b,c,d...)

I think ... is what you are looking for

arrays are mutable so in your matrix operation you are likely modifying the original arrays... it sounds like you just want to return a new dataset without modifying the existing x1..x10 arrays

of coarse there is probably a 75% chance I didnt understand what you were asking ...

answered Nov 1, 2013 at 15:51

Joran Beasley

114k13 gold badges167 silver badges187 bronze badges

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

4 Comments

Zhenquan Hu Over a year ago

oops! I missed something important: A is a 4-dimensional array, B is a 3-dimensional array. I could use myfunc() with 3 loops to generate B from A. But I wonder whether there is a method to reduce the dimension of A from 4 to 3 by calling myfunc()

Joran Beasley Over a year ago

post A and B ... but small size example ... I dont understand what your input is nor your expected output

Zhenquan Hu Over a year ago

Thanks a lot for your quick response! Just make it simpler: A is 3 dimension (5*5*10), B is 2 dimension (5*5), myfunc() is a multiple linear regression function which returns a single value from 10 input values. So for each A[i,j,:] could generate a B[i,j] by calling myfunc() using iterative method: for i in range(5): / for j in range(5): / B[i,j] = myfunc(A[i,j,:]). But is it possible to call myfunc() in matrix without loops?

Zhenquan Hu Over a year ago

For example: A = np.random.randn(5,5,10), C = A.sum(axis = 2) would return a 5*5 array, how to do a similar way to let B = A.myfunc(axis=2)

Collectives™ on Stack Overflow

Reduce dimension of numpy array by defined function

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related