CODEWITHSUNDEEP
pythonnumpy
Sun Bear
Sun Bear

Reputation: 8234

How to replace certain elements of a NumPy array via an index array

I have an numpy array a that I would like to replace some elements. I have the value of the new elements in a tuple/numpy array and the indexes of the elements of a that needs to be replaced in another tuple/numpy array. Below is an example of using python to do what I want.How do I do this efficiently in NumPy?

Example script:

a = np.arange(10)
print( f'a = {a}' )
newvalues = (10, 20, 35)
indexes = (2, 4, 6)
for n,i in enumerate( indexes ):
    a[i]=newvalues[n]
print( f'a = {a}' )

Output:

a =  array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
a =  array([ 0,  1, 10,  3, 20,  5, 35,  7,  8,  9])

I tried a[indexes]=newvalues but got IndexError: too many indices for array: array is 1-dimensional, but 3 were indexed

Upvotes: 0

Views: 798

Answers (1)

luuk
luuk

Reputation: 1855

The list of indices indicating which elements you want to replace should be a Python list (or similar type), not a tuple. Different items in the selection tuple indicate that they should be selected from different axis dimensions.

Therefore, a[(2, 4, 6)] is the same as a[2, 4, 6], which is interpreted as the value at index 2 in the first dimension, index 4 in the second dimension, and index 6 in the third dimension.

The following code works correctly:

indexes = [2, 4, 6]
a[indexes] = newvalues

See also the page on Indexing from the numpy documentation, specifically the second 'Note' block in the introduction as well as the first 'Warning' under Advanced Indexing:

In Python, x[(exp1, exp2, ..., expN)] is equivalent to x[exp1, exp2, ..., expN]; the latter is just syntactic sugar for the former.

The definition of advanced indexing means that x[(1,2,3),] is fundamentally different than x[(1,2,3)]. The latter is equivalent to x[1,2,3] which will trigger basic selection while the former will trigger advanced indexing. Be sure to understand why this occurs.

Upvotes: 3

Related Questions