Broadcasting - use Technical Quotient

3. Apply

Given arrays a = np.array([0, 10, 20, 30]) and b = np.array([1, 2, 3]), how would you reshape a and add it to b to use broadcasting?

a.reshape((1, 4)) + b.reshape((3, 1))

Incorrect; this reshapes both a and b unnecessarily and in a way that doesn't align with broadcasting rules.

a.reshape((4, 1)) + b

Correct; reshapes a to a (4, 1) array and adds it to b, leveraging broadcasting to get a (4, 3) result.

a + b.reshape((1, 3))

Incorrect; reshaping b is not necessary and does not affect the compatibility of a and b for broadcasting.

a[:, np.newaxis] + b[np.newaxis, :]

Incorrect; introduces a new axis for both a and b, but overcomplicates the operation compared to the direct a[:, np.newaxis] + b.

4. Analyze

You attempt to add arrays a = np.array([[0, 10, 20], [30, 40, 50]]) and b = np.array([0, 1, 2, 3]) in your financial analysis software, expecting to adjust each row in a by the corresponding values in b. However, you encounter a ValueError related to shapes not being broadcastable. What is the likely cause of this error, considering broadcasting rules?

The b array should have been reshaped to match a's dimensions exactly.

Incorrect; reshaping b to match a's dimensions could solve the problem but does not align with the fundamental issue of broadcasting compatibility.

Both a and b must be converted to 2D arrays with matching shapes.
You need to transpose a before adding it to b.

Incorrect; transposing a changes its shape but does not resolve the fundamental mismatch with b's shape for broadcasting.

Mismatched dimensions in a and b prevent direct addition via broadcasting.

Correct; the trailing dimensions of a (3) and b (4) do not match and are not compatible for broadcasting, leading to a ValueError. Broadcasting rules require that the trailing dimensions either match or one of them is 1.

1. Remember

2. Understand

3. Apply

4. Analyze

Leave a ReplyCancel Reply