Get costs at any given iteration

[daffidwilde]

I want to visualise the convergence rate of k-modes (and k-prototypes, eventually) by plotting the cost function at each iteration. I've had a look through kmodes.py and I can't see a way of extracting that information directly. Could you suggest a way of going about this?

Thanks,

Henry

[nicodv]

This is not currently supported, but should be an easy fix.

See here: https://github.com/nicodv/kmodes/blob/master/kmodes/kmodes.py#L274

In stead of just storing all_costs[best], we could also store all_costs.

Feel free to make a PR.

[daffidwilde]

What I actually mean is that I want to visualise the cost at each epoch within a single run of the algorithm, not just the final cost. Apologies if I am mistaken.

[daffidwilde]

Fixed as of #79.