finding rows of an array that all columns has inf value-CodePudding

I have an array that some rows have inf value,but I need to find rows that all columns have inf value. I can find one row with inf value but I need that all values of this row be inf. my array is nxm and all m column should have inf to be selected. do you know how can I do this in python? Thanks.

CodePudding user response：

Change equality if need all rows with "inf", written this way for better understanding

What you need to do find the minimum value in each row & check if that is not 'inf' than that row is eligible for output. Below is the example code:

import numpy as np
a = np.array([[np.inf,np.inf,np.inf,np.inf],[5,7,6,9],[1, 2, np.inf, 4]])
a[a.min(axis=1) != np.inf]

Output will be : array([[ 5., 7., 6., 9.], [ 1., 2., inf, 4.]])

Updating as per the OP question in comments - how to get row index of filtered row

np.where([a.min(axis=1) != np.inf])[1]

The output of this will be array([1, 2]), rows with atleast one non 'inf' value (in above example 2 & 3rd row, index 1 & 2).

Similarly

np.where([a.min(axis=1) == np.inf])[1]

This will return array([0]) , index of row (in above example case, 0th row) with all the 'inf' values.