[BUG] Z-normalisation numba function produces wrong results with integer inputs

[baraline]

Describe the bug

Functions such as z_normalise_series_2d from utils\numba\general.py when used with integer array as inputs ouput an integer array which do not correspond to expected normalization results. The cause is that these functions are using arr = np.zeros_like(X) to create the returned array, which will use the same dtype as the input X.

Replacing this with arr = np.zeros(X.shape) should fix the issue.

Steps/Code to reproduce the bug

from aeon.utils.numba.general import z_normalise_series_2d
z_normalise_series_2d(np.array([[1,2,3]],dtype=int))

Expected results

Out[64]: array([[-1.22474487, 0. , 1.22474487]])

Actual results

Out[65]: array([[-1, 0, 1]])

Versions

No response

[aryan0931]

hello , I would like to work on this issue.

[baraline]

Hey, sorry, just seen your message, this issue should be fixed with #2176.