Open
Conversation
…nd fitting methods
- Implement rigorous tests for Bessel K_v, inverse incomplete gamma functions, and Student's t CDF in `test_special.py`. - Validate against scipy.special and scipy.stats with tight tolerances to catch known errors. - Introduce tests for sample skewness and kurtosis in `test_stats.py`, ensuring consistency with scipy.stats. - Create extensive tests for all 11 univariate distributions in `test_univariate.py`, covering logpdf, CDF, stats, fitting, and edge cases. - Include checks for PDF integration, inverse consistency, parameter recovery, and gradient correctness. - Address specific findings related to Student-T and IG distributions to ensure accuracy and reliability.
…m_hat and v_hat calculations
…or accurate statistics
…d accuracy using log_kv
…ccurate statistics
…on and improve stability using log_kv
…mproved performance and stability
… precision and stability; enhance initial approximation and Newton-Halley refinement process.
…ts for extreme values and refine tolerance levels for better precision.
…e using method of moments
… handling; add explicit normalization for total PDF integral to ensure accurate results.
…initial estimates; replace kv with log_kv for stability and add sample moments computation.
…g_kv for stability, add expected value computations, and implement negative log-likelihood gradient calculation.
…likelihood computation and update fit method to support EM approach.
…ctor _fit_mle and _fit_ldmle to use sample moments for better convergence.
… _sample_moments method and update _fit_mle to use these estimates for improved fitting.
…re functions for improved parameter estimation.
…special for accuracy and properties.
… enhance fit method to support MOM and MLE options.
… sampling accuracy
…igns and enforce CDF monotonicity for improved accuracy
…documentation for small-lambda thresholds in Kolmogorov-Smirnov and Cramér-von Mises tests
…ribution within int64 bounds
…nd avoid division by zero
…re for distribution categories
…ase handling and improving parameterization for distributions.
…; optimize matrix calculations and enhance method validation.
…d in Correlation class
…g for valid correlation output; update laloux method to use correlation from covariance.
…e descriptions of methods and parameters for clarity.
…ce functions; add autofunction directives for better clarity and accessibility.
…d MvtStudentT classes to include optional x parameter for improved flexibility in LD-MLE optimization.
… for LD-MLE optimization
…t ECME fitting algorithm for improved parameter estimation
…a=0 consistency with MVT Student-T, and ECME fitting validation
- Removed unused methods for reconstructing copula parameters in MvtGH, MvtNormal, MvtSkewedT, and MvtStudentT classes. - Improved PPF computation in _ppf.py by implementing a unified dispatcher for cubic interpolation and Brent root-finding with custom VJP rules. - Added shared IFT backward pass for PPF to ensure efficient gradients. - Enhanced cubic PPF function creation with captured parameters for better flexibility.
- consolidate duplicate fitting code - implement persistent adam state + standardise grad clipping - add early stopping - convert inner python for loop to lax.scan
…integrate them into Student-T, Skewed-T and GH copula parameter estimation
…-T nu estimation to use bisection method
…ialization and update prev_ll to -1e20 for stability
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
test_special.py.test_stats.py, ensuring consistency with scipy.stats.test_univariate.py, covering logpdf, CDF, stats, fitting, and edge cases.