Changelog#

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[v1.0.1] - 2025-12-03#

b-significant test for evaluating spatial and temporal significance of b-value variations, including plotting functionality and comprehensive documentation.
Lilliefors test implementation for goodness-of-fit testing of the Gutenberg-Richter distribution.
Bootstrap method to BValue class for uncertainty estimation of b-value calculations.
Weights support for Kolmogorov-Smirnov test (ks_test_gr) and frequency-magnitude distribution plotting functions to handle weighted catalogs.
value property for both AValue and BValue estimator classes for consistent access to estimated values.
std property for AValue estimator class to access standard deviation of a-value estimates.
is_estimated property and n property to both AValue and BValue classes for better consistency between the two estimator classes.
Safety mechanism for values_from_partitioning to handle edge cases more robustly.
get_options function to module initialization for accessing configuration options.
Comprehensive test coverage for a-value estimators, b-significant test, Lilliefors test, and Kolmogorov-Smirnov test.

Same as get_cum_fmd, plot_cum_fmd (both function and Catalog method) now take fmd_bin as input instead of delta_m, to reflect that this argument doesn’t necessarily match the discretization of magnitudes, and be consistent with get_fmd and plot_fmd which already have this modification.
Optimized maximum magnitude finding in estimate_mc functions for improved performance.
Optimized simulation of truncated exponential distribution for better computational efficiency.
Restricted bootstrap method to only work within the b-more-positive estimation method where it is statistically applicable.
Improved robustness of a-value estimation to handle edge cases better.
Adjusted Lilliefors test (p_lillierfors) to be compatible with the b-positive estimation method.
Bootstrap standard deviation calculation now orders magnitudes by time and removes NaN values for more accurate uncertainty estimates.
Improved colors and labels in plotting functions for better visualization and consistency.
Adjusted plotting functions to produce fewer warnings during normal usage.
Reorganized example notebooks: removed redundant catalog notebook and cleaned up the how-to notebook for better clarity.
User documentation and API reference extensively updated to reflect new features and improvements.

Binning precision issues by introducing epsilon tolerance to avoid incorrect magnitude binning due to floating-point arithmetic.
Rounding behavior in binning functions that was inconsistent with numpy’s round() behavior, ensuring consistent bin assignment.
Error message for edge case when delta_m < epsilon to provide clearer feedback.
Bug in moransI spatial analysis function.
Axis input handling for seismicity plotting functions to work correctly with custom axes.
Minor error in bootstrap calculation that could affect uncertainty estimates.
Various typos in documentation and docstrings.

Add your changes to the relevant section below, while uncommenting the section header.

API and User Documentation has been extensively updated and improved.
Added methods to calculate statistics on the RateGrid class and concatenate multiple RateGrid objects.

Small changes to the plots module api to make it more consistent.
Updated the example jupyter notebooks to reflect the new API.
estimate_mc and plot_fmd take now fmd_bin as input instead of delta_m, to reflect that this is independent from the discretization of the catalog magnitudes.
Functions and methods to estimate the mc value now return a consistent format. Instead of a tuple with a variable amount of values, they now return a tuple with first the best_mc value and then a dictionary with the additional values.

In the binning_test, shift the array to the smallest value, to avoid that the bin-center has an effect on the test.

API reference for a and b value packages, including full list of attributes.
Thorough API documentation for the Catalog class.
Added estimate_mc_maxc and estimate_mc_b_stability methods to the Catalog class.
Option to download catalogs in batches from FDSNWS in order to avoid timeouts.

analysis.ks_test_gr now takes the b-value as a parameter instead of beta.
Catalog.estimate_mc got updated to the latest version of the mc_ks function and renamed to Catalog.estimate_mc_ks.
Improve and streamline the estimate_mc functions and slightly change their input parameters.
Catalog.estimate_b got updated to the class-based API of the b_value package.
Catalog.estimate_a added to the Catalog class, including full documentation.