Suggest adding float ranges in binning_table.build result

[chengs]

In such code

optb = OptimalBinning(name='example', dtype='numerical', solver='cp')
optb.fit(X, y)
binning_table = optb.binning_table
bins = binning_table.build()
print(bins)

binning_table.build method returns a pandas.DataFrame as follows:

This dataframe actually can be used more than display, with other format, such as JSON.
For example, with this table, one can build a transform function converting values into WoE, in other language/platform (C, tensorflow, java etc)

Therefore, I would suggest that, the range of each bin can be saved in their raw float values, and then this build api can be modified to accept a format property (or add a new build_json api).

If this is reasonable to you, I am happy to make a PR.

[guillermo-navas-palencia]

Hi @chengs. The suggested feature might be already covered by #96. See https://gnpalencia.org/optbinning/binning_binary.html#optbinning.OptimalBinning.to_json. Would this suffice?