Change some of numpy array columns according to logical condition

Question

I have a 2D NumPy array, and I want to perform the following operation:

For each column in the array, which is a series of non-decreasing values, replace this column with a column of the differences (that is, each entry is the difference between the two previous ones).

Every other column remains the same (except that first row is removed to fit to the differences columns dimension).

For example, in the matrix:

[ [1,1,1,2,3,4]
  [1,3,4,3,4,5]
  [1,7,3,4,2,7] ]

The differences matrix is:

[ [0,2,3,1,1,1]
  [0,4,-1,1,-1,2] ]

And thus the third and fifth columns which have decreasing values will remain the same, while other columns are replaced with the differences columns, resulting with:

[ [0,2,4,1,4,1]
  [0,4,3,1,2,2] ]

I tried something like this:

tempX = np.diff(X, axis = 0).transpose()
return np.where(tempX >= 0, tempX, X[1:].transpose())

But the condition in np.where is performed element-wise and not for each column (or row).

How can I change the condition so it will work? Is there a more efficient way to implement this?

Could you use a bigger array as sample and more variety in numbers? — Divakar
– Divakar, Commented Nov 15, 2016 at 10:31
It's hard to understand if there is a typo in your example. ;) I guess, it should be "will be changed to [ [0, 2, 1] [0, 4, 3] ]", right? — nostradamus
– nostradamus, Commented Nov 15, 2016 at 10:33

quantummind · Accepted Answer · 2016-11-15 11:06:56Z

1

You can try it so:

b = a[1:] - a[:-1]
decrease = numpy.where(numpy.min(b, axis=0)<0)
b[:,decrease] = a[1:, decrease]

You can also do that in one expression:

numpy.where(numpy.min(a[1:]-a[:-1],axis=0)>=0, a[1:]-a[:-1], a[1:])

edited Nov 15, 2016 at 11:06

answered Nov 15, 2016 at 10:47

quantummind

2,1461 gold badge15 silver badges20 bronze badges

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

Comments

Divakar · Accepted Answer · 2016-11-15 10:44:12Z

0

You could use boolean-indexing -

# Get the differentiation along first axis
diffs = np.diff(a,axis=0)

# Mask of invalid ones
mask = (diffs<0).any(0)

# Use the mask to set the invalid ones to the original elements
diffs[:,mask] = a[1:,mask]

Sample run -

In [141]: a
Out[141]: 
array([[1, 1, 1, 2, 3, 4],
       [1, 3, 4, 3, 4, 5],
       [1, 7, 3, 4, 2, 7]])

In [142]: diffs = np.diff(a,axis=0)
     ...: mask = (diffs<0).any(0)
     ...: diffs[:,mask] = a[1:,mask]
     ...: 

In [143]: diffs
Out[143]: 
array([[0, 2, 4, 1, 4, 1],
       [0, 4, 3, 1, 2, 2]])

answered Nov 15, 2016 at 10:44

Divakar

222k19 gold badges273 silver badges374 bronze badges

Collectives™ on Stack Overflow

Change some of numpy array columns according to logical condition

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related