I have a list of arrays, and I want to pass the first 10 columns of each array into a scaler to transform them, but not the rest of the columns as they are dummy variables.
Each individual array is 2D and contains data corresponding to a specific column.
I have tried:
list[:][:10]
But this gives me simply the first 10 arrays, rather than all of the arrays' first 10 columns.
You can convert the outer list into a numpy array too (np.array(my_list))
and then use multidimensional indexing like: my_np_list[:, :10]
Assuming that each array has the same length, try np.stack with indexing -
arr = [np.array([1,2,3,4]),
np.array([4,5,6,7]),
np.array([8,9,10,11])]
#Getting the first 2 column (:10 for first 10)
np.stack(arr)[:,:2]
array([[1, 2],
[4, 5],
[8, 9]])
list[:]is just a slice of the list, which will give you the whole list back and then [:10] will give the first 10 elements, as observed. If you can't stack all arrays into a 3D array, you'll have to iterate through the list, i.e.,[arr[:10] for arr in list].