feat: add to_dict method for binning table serialization

[sudo-hannes]

Add to_dict method for binning table serialization

Summary

Implements a to_dict() method that converts optimal bins, split points, and transformations to dictionary format for easy serialisation and export.

[sudo-hannes]

@guillermo-navas-palencia

I saw that the test failed because "http://lib.stat.cmu.edu/datasets/boston" is no longer available. I found a similar dataset on Kaggle. When I use this dataset, all but one of the tests pass.

Can I open a separate PR to fix the test issue?

optbinning/tests/datasets/datasets.py

Lines 1 to 20 in 0720b8d

	import numpy as np
	import pandas as pd
	class Data:
	def __init__(self, data, target, feature_names):
	self.data = data
	self.target = target
	self.feature_names = feature_names
	def load_boston():
	data_url = "http://lib.stat.cmu.edu/datasets/boston"
	raw_df = pd.read_csv(data_url, sep=r"\s+", skiprows=22, header=None)
	raw_data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
	target = raw_df.values[1::2, 2]
	feature_names = ['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS',
	'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT']
	return Data(raw_data, target, feature_names)

[guillermo-navas-palencia]

@guillermo-navas-palencia

I saw that the test failed because "http://lib.stat.cmu.edu/datasets/boston" is no longer available. I found a similar dataset on Kaggle. When I use this dataset, all but one of the tests pass.

Can I open a separate PR to fix the test issue?

optbinning/tests/datasets/datasets.py

Lines 1 to 20 in 0720b8d

	import numpy as np
	import pandas as pd
	class Data:
	def __init__(self, data, target, feature_names):
	self.data = data
	self.target = target
	self.feature_names = feature_names
	def load_boston():
	data_url = "http://lib.stat.cmu.edu/datasets/boston"
	raw_df = pd.read_csv(data_url, sep=r"\s+", skiprows=22, header=None)
	raw_data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
	target = raw_df.values[1::2, 2]
	feature_names = ['CRIM', 'ZN', 'INDUS', 'CHAS', 'NOX', 'RM', 'AGE', 'DIS',
	'RAD', 'TAX', 'PTRATIO', 'B', 'LSTAT']
	return Data(raw_data, target, feature_names)

Hi @sudo-hannes. Thanks for your contribution! Please feel free to work on that PR :). In addition, we could consider saving a csv file and loading it directly without relying on external sources.