Add Manhattan and Minkowski distances to KNN with doctests

[kdt523]

This PR extends the K-Nearest Neighbours (KNN) implementation to support additional distance metrics while keeping Euclidean as the default.
What’s included:
Adds distance_metric parameter to KNN: 'euclidean' (default), 'manhattan', 'minkowski'
Adds p parameter for Minkowski (Lp) distance (validated p >= 1)
Implements _manhattan_distance and _minkowski_distance helpers
Updates classify() to use the selected metric
Stores training data as a list to avoid zip exhaustion on multiple calls
Moves the scikit-learn demo under main so importing the module has no extra dependency
Doctests: distances, classification across metrics, and error handling
Backward compatibility:
Existing code remains unchanged by default
(distance_metric='euclidean').
No documentation changes.
How to verify locally:
From the Python folder:
Doctests:
python -m doctest -v [k_nearest_neighbours.py]
Fixes #13546

[kdt523]

Hi @TheAlgorithms/maintainers 👋
This PR Add Manhattan and Minkowski distances to KNN with doctests #13555
Could someone with write access please review and approve it when you get a chance?
Thanks a lot for your time and help!

[kdt523]

Hi @TheAlgorithms/maintainers
This PR #13555 Add Manhattan and Minkowski distances to KNN with doctests
Could someone with write access please review and approve it when you get a chance?
Thanks a lot for your time and help!