44

I'm trying to return maximum values of multiple array in an element-wise comparison. For example:

A = array([0, 1, 2]) B = array([1, 0, 3]) C = array([3, 0, 4])

I want the resulting array to be array([3,1,4]).

I wanted to use numpy.maximum, but it is only good for two arrays. Is there a simple function for more than two arrays?

Improve this question

3 Answers 3

Reset to default
67

With this setup:

>>> A = np.array([0,1,2]) >>> B = np.array([1,0,3]) >>> C = np.array([3,0,4])

You can either do:

>>> np.maximum.reduce([A,B,C]) array([3, 1, 4])

Or:

>>> np.vstack([A,B,C]).max(axis=0) array([3, 1, 4])

I would go with the first option.

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

3 Comments

I would try to hold the data in a two-dimensional array right from the start, and then use the second option.
@SvenMarnach The axis argument calls reduce in the end, I don't understand why it matters.
Note that unlike np.maximum, np.maximum.reduce won't work out of the box with masked array. Neither is np.ma.maximum.reduce(…, axis=0) . You can manually keep track of the masks however.
22

You can use reduce. It repeatedly applies a binary function to a list of values...

For A, B and C given in question...

np.maximum.reduce([A,B,C]) array([3,1,4])

It first computes the np.maximum of A and B and then computes the np.maximum of (np.maximum of A and B) and C.

Improve this answer

Comments

1

You can also use:

np.column_stack([A, B, C]).max(axis=1)

The result is the same as the solutions from the other answers.

I use Pandas more heavily than NumPy so for me it's easier to think of 1D arrays as something similar to Pandas Series. The above would be equivalent to:

pd.concat([A, B, C], axis=1).max(axis=1)
Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.