I want to convert this sum(sum(W_set{1}~=0)) into Python:
%MATLAB code:
W_set = {[[[-0.05747274, -0.05268928],
[-0.10961724, 0.10119643],
[-0.0327577 , 0.01514941]]],
[[[-0.05557293, 0.11311244],
[-0.22935626, -0.11837874],
[-0.05567432, -0.0558801 ]]]};
sum(sum(W_set{1}~=0))
The output is 6. how can I get this output in Python, W_set in Python is a list.
I have tried the following approach, but the result is not the same.
#Python code:
import numpy as np
W_set = [[[[-0.05747274, -0.05268928],
[-0.10961724, 0.10119643],
[-0.0327577 , 0.01514941]]],
[[[-0.05557293, 0.11311244],
[-0.22935626, -0.11837874],
[-0.05567432, -0.0558801 ]]]]
if np.array(W_set[0]).all != 0: #checking not equal to zero
sum_ = np.sum(np.sum(W_set[0])) #sum
print(sum_)
The answer is -0.13619112. What is the meaning of operator ~= when used in the MATLAB sum function?
1 Answer
You can use numpy.sum it will sum over all the axis, but in order to compare the elements to zero in a vectorized instruction you need it to be a numpy array
np.sum(np.array(W_set[0]) != 0)
Alternatively you could have made W_set = np.array([...]) and then
np.sum(W_set[0] != 0)
(W_set[0] != 0).sum()
