Runtime configuration for statistical propagation, fitting and nuisance handling. More...
#include <StatisticManager.h>
Public Attributes | |
| std::size_t | MC_draws = 100 |
| Number of accepted MC draws used for uncertainty propagation. | |
| std::size_t | MC_threads = 1 |
| Number of worker threads used by MC propagation. | |
| unsigned int | MC_seed = 123456u |
| RNG seed used for reproducible MC nuisance and experimental-data sampling. | |
| double | skew_abs_threshold = 0.2 |
| Absolute skewness threshold below which a summary is treated as symmetric. | |
| bool | print_mc_progress = true |
| Print MC progress with ETA based on measured draw time. | |
| bool | print_chi2_pipeline_progress = false |
| Print chi-square workflow stages after/beside the MC progress bar. | |
| bool | print_mc_config = false |
| Print nuisance candidates and retained MC marginal configuration. | |
| bool | print_fit_summary = false |
| Print high-level fit backend summaries. | |
| bool | print_scan_summary = false |
| Print likelihood-scan summaries. | |
| bool | print_cache_summary = false |
| Print internal cache diagnostics. | |
| bool | print_debug = false |
| Master debug flag for low-level diagnostic output. | |
| bool | write_mc_samples_csv = false |
| Write accepted MC observable samples to CSV. | |
| std::string | mc_samples_csv_path = "obs_samples.csv" |
| Output CSV path used when write_mc_samples_csv is true. | |
| std::size_t | mc_progress_probe_draws = 5 |
| Number of first accepted draws used to stabilize the first ETA. | |
| std::size_t | mc_progress_update_every = 1 |
| Accepted-draw stride between progress updates. | |
| std::shared_ptr< StatisticProgressMonitor > | progress_monitor {} |
| Optional thread-safe progress sink for GUI/notebook frontends. | |
| std::map< ParamId, std::pair< double, double > > | fit_parameter_bounds {} |
| Optional explicit minimizer bounds keyed by fit ParamId. | |
| std::map< ParamId, double > | fit_parameter_offsets {} |
| Optional affine display offsets: model value = fitted value - offset. | |
| AdvancedStatisticConfig | advanced {} |
| Advanced fit/pruning/covariance configuration. | |
Detailed Description
Runtime configuration for statistical propagation, fitting and nuisance handling.
Basic runtime configuration for statistical propagation.
The defaults are intentionally conservative: a Gaussian copula is used for both nuisance and experimental distributions, HESSE is enabled for the MLE, and local sensitivity pruning is enabled to reduce the nuisance set before expensive fits.
Keep this structure small and user-facing: it contains the knobs that are commonly changed in scripts and examples. Advanced fit/pruning/covariance controls live in AdvancedStatisticConfig and are grouped under advanced to avoid an overloaded top-level config object.
Definition at line 121 of file StatisticManager.h.
Member Data Documentation
◆ advanced
| AdvancedStatisticConfig StatisticConfig::advanced {} |
Advanced fit/pruning/covariance configuration.
Definition at line 145 of file StatisticManager.h.
◆ fit_parameter_bounds
| std::map<ParamId, std::pair<double, double> > StatisticConfig::fit_parameter_bounds {} |
Optional explicit minimizer bounds keyed by fit ParamId.
Definition at line 142 of file StatisticManager.h.
◆ fit_parameter_offsets
| std::map<ParamId, double> StatisticConfig::fit_parameter_offsets {} |
Optional affine display offsets: model value = fitted value - offset.
Definition at line 143 of file StatisticManager.h.
◆ MC_draws
| std::size_t StatisticConfig::MC_draws = 100 |
Number of accepted MC draws used for uncertainty propagation.
Definition at line 122 of file StatisticManager.h.
◆ mc_progress_probe_draws
| std::size_t StatisticConfig::mc_progress_probe_draws = 5 |
Number of first accepted draws used to stabilize the first ETA.
Definition at line 137 of file StatisticManager.h.
◆ mc_progress_update_every
| std::size_t StatisticConfig::mc_progress_update_every = 1 |
Accepted-draw stride between progress updates.
Definition at line 138 of file StatisticManager.h.
◆ mc_samples_csv_path
| std::string StatisticConfig::mc_samples_csv_path = "obs_samples.csv" |
Output CSV path used when write_mc_samples_csv is true.
Definition at line 136 of file StatisticManager.h.
◆ MC_seed
| unsigned int StatisticConfig::MC_seed = 123456u |
RNG seed used for reproducible MC nuisance and experimental-data sampling.
Definition at line 124 of file StatisticManager.h.
◆ MC_threads
| std::size_t StatisticConfig::MC_threads = 1 |
Number of worker threads used by MC propagation.
Definition at line 123 of file StatisticManager.h.
◆ print_cache_summary
| bool StatisticConfig::print_cache_summary = false |
Print internal cache diagnostics.
Definition at line 132 of file StatisticManager.h.
◆ print_chi2_pipeline_progress
| bool StatisticConfig::print_chi2_pipeline_progress = false |
Print chi-square workflow stages after/beside the MC progress bar.
Definition at line 128 of file StatisticManager.h.
◆ print_debug
| bool StatisticConfig::print_debug = false |
Master debug flag for low-level diagnostic output.
Definition at line 133 of file StatisticManager.h.
◆ print_fit_summary
| bool StatisticConfig::print_fit_summary = false |
Print high-level fit backend summaries.
Definition at line 130 of file StatisticManager.h.
◆ print_mc_config
| bool StatisticConfig::print_mc_config = false |
Print nuisance candidates and retained MC marginal configuration.
Definition at line 129 of file StatisticManager.h.
◆ print_mc_progress
| bool StatisticConfig::print_mc_progress = true |
Print MC progress with ETA based on measured draw time.
Definition at line 127 of file StatisticManager.h.
◆ print_scan_summary
| bool StatisticConfig::print_scan_summary = false |
Print likelihood-scan summaries.
Definition at line 131 of file StatisticManager.h.
◆ progress_monitor
| std::shared_ptr<StatisticProgressMonitor> StatisticConfig::progress_monitor {} |
Optional thread-safe progress sink for GUI/notebook frontends.
Definition at line 140 of file StatisticManager.h.
◆ skew_abs_threshold
| double StatisticConfig::skew_abs_threshold = 0.2 |
Absolute skewness threshold below which a summary is treated as symmetric.
Definition at line 125 of file StatisticManager.h.
◆ write_mc_samples_csv
| bool StatisticConfig::write_mc_samples_csv = false |
Write accepted MC observable samples to CSV.
Definition at line 135 of file StatisticManager.h.
The documentation for this struct was generated from the following files:
- /home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/core/src/Statistic/domain/StatisticManager.h
- /home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/pyhyperiso/core/Statistic/StatisticConfig.py
