API: follow sklearn api conventions

[martinfleis]

Closes #55

• avoid modification of parameters
• move checks to fit()
• store fitted parameters with an underscore
• ensure fit() has a sklearn-compatible signature
• return self in fit()
• ensure new call to fit() clears the data

Generally following https://scikit-learn.org/stable/developers/develop.html when it makes sense to follow it.

Other plotting-classes within the scikit-learn package typically implement “from_estimator” and “from_predictions” class methods (e.g. ConfusionMatrixDisplay, PredictionErrorDisplay, etc.). In this way it is easy to use the class with any class that would provide clustering data. The current implementation of the class provides from_data() and from_centers() methods, which achieve similar, but not directly compatible behavior.

Given the nature of Clustergram being typically composed of a series of estimators differing in a number of clusters, it is not possible to create an API like this that is compatible with existing sklearn plotting classes. I suppose that our existing alternative constructors are sufficient in this case.

[codecov-commenter]

Codecov Report

Patch coverage: 96.10% and project coverage change: +1.62 🎉

Comparison is base (b64916c) 80.86% compared to head (8b4d44c) 82.49%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #59      +/-   ##
==========================================
+ Coverage   80.86%   82.49%   +1.62%     
==========================================
  Files           2        2              
  Lines         345      377      +32     
==========================================
+ Hits          279      311      +32     
  Misses         66       66              

Impacted Files	Coverage Δ
clustergram/clustergram.py	82.16% <96.10%> (+1.68%)	⬆️

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.