pythonnp.rollaxis

The numpy rollaxis() function is used to roll the specified axis backwards, until it lies in a specified position. This function is very useful when working with multi-dimensional arrays, as it allows you to change the shape of the array without changing the data.

The function takes three arguments:

numpy.rollaxis(arr, axis, start)

Here,

- arr: Input array.
- axis: The axis to roll backwards. It can be any integer or tuple of integers. If a tuple is specified, the first element is the axis to roll, and the second element is the new position of the rolled axis.
- start: The new position of the rolled axis. It can be any integer.

The function returns a new array with the specified axis rolled to the new position.

Example:

Consider the following 3D array:

arr = np.ones((3, 4, 5))

To roll the second axis to the end:

np.rollaxis(arr, 1, 3)

This will return a new array with shape (3, 5, 4), where the second axis has been rolled to the end.