How to set random values to 2d-numpy-array where values are very low?

Question

I have a 2d-numpy-array (28x28) and would like to set random values to all that indices which have a value lower than 50. I tried something:

myarray[myarray < 50] = random.randint(80,100)

This doesn't work correctly because all selected indices are getting the same random value. I would like to have different random values.

for y in range(28):
   for x in range(28):
      if myarray[y,x] < 50:
         myarray[y,x] = random.randint(80,100)

The code above is solving the problem, but I think that it doesn't work efficently, because it gets very slow with big datasets.

Is there any better function?

You can create random arrays of a specified size e.g numpy.random.randint(80,100,(3,2)) would create array of shape [3,2] , so just create a random array the size of myarray<50, and put that instead docs.scipy.org/doc/numpy/reference/generated/… — user3376851
– user3376851, Commented Apr 18, 2019 at 10:16

FindOutIslamNow · Accepted Answer · 2019-04-18 10:10:54Z

1

Then,

myarray[myarray < 50] = np.random.uniform(low=80, high=100,
       size=myarray[myarray < 50].shape)

answered Apr 18, 2019 at 10:04

FindOutIslamNow

1,2661 gold badge16 silver badges37 bronze badges

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

1 Answer 1