Row assignment to numpy array

Question

>>> arr = np.array([[3, -4, 4], [1, -2, 2]])
>>> arr
array([[ 3, -4,  4],
       [ 1, -2,  2]])
>>> arr[1]-(1/3)*arr[0]
array([ 0.        , -0.66666667,  0.66666667])
>>> arr[1] = arr[1]-(1/3)*arr[0]
>>> arr
array([[ 3, -4,  4],
       [ 0,  0,  0]])

what do I do wrong? I want to assign the result of the calculation to the second row of the array "arr"

You can't assign back because its an int array, doesn't support float pt data. — Divakar
– Divakar, Commented Jun 26, 2017 at 14:18

willeM_ Van Onsem · Accepted Answer · 2017-06-26 14:19:43Z

2

The problem is that you constructed an array of ints. You can construct an array of np.floats by using the dtype parameter:

arr = np.array([[3, -4, 4], [1, -2, 2]],dtype=np.float)

Python has a dynamic approach when it comes to types: every element in a list can have a different type. But numpy works with matrices where all elements have the same type. Therefore assigning a float to an int matrix, will convert the row first to ints.

This will construct an array:

>>> arr = np.array([[3, -4, 4], [1, -2, 2]],dtype=np.float)
>>> arr[1] = arr[1]-(1/3)*arr[0]
>>> arr
array([[ 3.        , -4.        ,  4.        ],
       [ 0.        , -0.66666667,  0.66666667]])

answered Jun 26, 2017 at 14:19

willeM_ Van Onsem

482k33 gold badges483 silver badges624 bronze badges

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

Collectives™ on Stack Overflow

Row assignment to numpy array

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related