Standardize Then Train

Standardise every column of a training matrix: subtract the column mean and divide by the column population standard deviation (ddof=0). A constant column maps to zeros. Only the supplied matrix is used, so no information from a test set can leak in.

Implement solve(x).

Returns. Return a float NumPy array with the same shape as x, where each non-constant column has mean $00$ and standard deviation $11$.

Examples

Example 1

Input:

solve([[1, 10], [3, 14], [5, 18]])

Output:

[[-1.224745, -1.224745], [0.0, 0.0], [1.224745, 1.224745]]

Example 2

Input:

solve([[2, 7], [2, 9]])

Output:

[[0.0, -1.0], [0.0, 1.0]]