Implement the Bayesian Blocks optimal histogram binning algorithm to eliminate the only TODO in the codebase.

Background:
Currently, BinMethod::Bayesian falls back to Freedman-Diaconis rule. The proper implementation should use the Bayesian Blocks algorithm (Scargle et al., 2013) which uses dynamic programming to find optimal change points.

Algorithm:

• Input: Data points, prior on number of blocks
• Output: Optimal histogram bin edges
• Method: O(n²) dynamic programming
• Finds change points that maximize posterior probability

Acceptance Criteria:

• Implement Bayesian Blocks algorithm in stats module
• Add comprehensive tests (EXTREME TDD)
• Add example demonstrating the algorithm
• Update documentation
• Remove TODO comment from code

References:

• Scargle, J. D., et al. (2013). "Studies in Astronomical Time Series Analysis. VI. Bayesian Block Representations." ApJ, 764, 167.

Labels: enhancement, stats