loop over NxNxN array in python and replace all values matching a criterion

Question

I have an nxnxn matrix which I want to loop over and replace all values < 1E-35 with 1E-35.

          for i in range(N):
              for j in range(N):
                  for k in range(N):
                      if data[i][j][k] < 1E-35:
                          data[i][j][k] = 1E-35

Doesn't seem to work.

Edit: I worked it out. It was both the indentation and incorrect indexing [i][j][k].

Problem remains: this loops from 0 to N-1 of the NxNxN data? When I do data.min() I still get values ~ 1E-101 which should be 1E-35 after the loop. Am I doing the for loop wrong?

Perhaps a few more details. In what way does it not seem to work? Also you may want to fix the indentation of your if statement. What is the datatype of data? — Steven Rumbalski
– Steven Rumbalski, Commented Jan 16, 2013 at 19:05
Griff -- If you worked it out, post it as a solution (if others haven't already) and accept it when SO allows. That way, if someone else comes along with the same problem, they'll be able to benefit from what you learned. If someone has already posted a good answer (or even a better answer then you worked out yourself), feel free to accept. — mgilson
– mgilson, Commented Jan 16, 2013 at 19:15
Sorry, I have amended it - it just seems so basic and that it would have been asked somewhere else if I had looked harder. — Griff
– Griff, Commented Jan 16, 2013 at 19:18

mgilson · Accepted Answer · 2013-01-16 19:13:35Z

5

It looks to me like you are using numpy in which case, you may want np.where:

data = np.where(data < 1e-35,1e-35,data)

Alternatively, you could use fancy indexing:

data[ data < 1e-35 ] = 1e-35

edited Jan 16, 2013 at 19:13

answered Jan 16, 2013 at 19:08

mgilson

312k70 gold badges656 silver badges722 bronze badges

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

2 Comments

Ashwini Chaudhary Over a year ago

data[i,j,k] is valid in numpy?

mgilson Over a year ago

@AshwiniChaudhary -- Yep. You can even do fun slicing: data[1:6,3:2,4:8:2]. If you ever need work with arrays as first-class objects in python, numpy is AWESOME.

Collectives™ on Stack Overflow

loop over NxNxN array in python and replace all values matching a criterion

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related