summaryrefslogtreecommitdiff
path: root/lib/utils.py
AgeCommit message (Collapse)AuthorLines
2022-09-23analyze-kconfig: perform param-shift in aggregate rather than in observationsDaniel Friesel-17/+1
this re-uses a function that is already present for analyze-archive anyways
2022-09-23deprecate --boolean-parametersDaniel Friesel-0/+2
2022-09-23ignore_param: work on by_name rather than observationsDaniel Friesel-8/+20
2022-09-21observations_to_by_name: sanity check for parameter vectorsDaniel Friesel-0/+2
2022-09-20utils: add support for more efficient observations file formatDaniel Friesel-1/+13
2022-08-31utils: observations_ignore_param: handle unknown parametersDaniel Friesel-1/+9
2022-06-22analyze-kconfig: allow individual parameters to be ignoredDaniel Friesel-0/+6
2022-06-15shift_param_in_aggregate: add '*=...' supportDaniel Friesel-7/+15
2022-06-15param_to_ndarray: handle parameters with numeric and non-numeric valuesDaniel Friesel-1/+3
2022-06-14analyze-kconfig: add param-shift supportDaniel Friesel-0/+16
2022-06-03add first order linear functions for evaluation purposesDaniel Friesel-0/+51
2022-05-13move enum_to_bool convertion to libDaniel Friesel-3/+41
2022-01-31allow parameter values to be adjusted before modelingDaniel Friesel-0/+21
(e.g. to ensure they're positive)
2021-10-26kconfig loader: load kconfig from its base directoryDaniel Friesel-0/+12
this ensures that source statements are handled correctly
2021-09-30observations_to_by_name: extract attributes name from observationsDaniel Friesel-6/+17
2021-09-30observations_to_by_name: None values are illegalDaniel Friesel-0/+4
2021-08-19ParallelParamFit -> ParamFit, optionally without parallelismDaniel Friesel-0/+14
2021-06-25add observations_to_by_nameDaniel Friesel-0/+31
2021-03-17always handle co-dependent parametersDaniel Friesel-1/+5
2021-03-10hardcode arg_support_enabled = TrueDaniel Friesel-1/+0
2021-03-02Move ParallelParamFit to a separate moduleDaniel Friesel-0/+74
2021-02-18refactor model generation from Analytic/PTAModel into ModelAttribute classDaniel Friesel-0/+10
Iteration over states/transitions and model attributes is no longer hardcoded into most model generation code. This should make support for decision trees and sub-states much easier.
2021-01-12add z-score based outlier detection and optional removalDaniel Friesel-1/+41
Note that outliers are removed from both training and evaluation data
2020-11-03analyze number of substates per stateDaniel Friesel-0/+27
2020-11-02Do not pass the entirety of by_nane/by_param to subprocessesDaniel Friesel-3/+1
This should speed up analysis quite a bit and also reduce memory usage significantly
2020-11-02Merge branch 'master' into merge-prep/janisDaniel Friesel-0/+13
2020-10-30Add DFATOOL_EXPORT_LASYNC variable for ET+LA / ET+Timer sync evalDaniel Friesel-0/+13
2020-10-08meopDaniel Friesel-2/+3
2020-10-08BA Janis import: working detection of optimal number of changepointsDaniel Friesel-0/+2
2020-07-03more consistent logging; use logger. instead of logging. where appropriateDaniel Friesel-1/+3
2020-07-02Use logging module for debug outputDaniel Friesel-10/+0
2020-05-28use black(1) for uniform code formattingDaniel Friesel-29/+53
2020-04-29move running_mean helper from dfatool to utilsDaniel Friesel-0/+11
2019-12-11oopsDaniel Friesel-0/+7
2019-11-26autopep8Daniel Friesel-8/+21
2019-10-11add helper functions for co-dependent parameter detectionDaniel Friesel-0/+17
2019-10-07Move codependent parameter logic to ParamStats / parameters.pyDaniel Friesel-0/+49
2019-10-07move ParamStats and helper functions to lib/parameters.pyDaniel Friesel-253/+0
2019-10-02improve co-dependent parameter detection logicDaniel Friesel-17/+20
also makes distinct_param_values more deterministic
2019-10-01initial check for boolean parameters' influence on other parametersDaniel Friesel-3/+15
2019-09-30std_by_param: return stddev matrix by individual parameter valuesDaniel Friesel-27/+38
2019-09-27prepare stddev-by-param-value matrix for fine-grained parameter analysisDaniel Friesel-3/+48
2019-09-25is_numeric: Also accept float valuesDaniel Friesel-2/+2
2019-08-16documentationDaniel Friesel-12/+25
2019-08-16prune_dependent_parameters: do not calculate bogus corrcoef valuesDaniel Friesel-9/+11
2019-08-16optionally prune dependent parameters before analysisDaniel Friesel-0/+66
2019-08-14functions, utils: pydocstyleDaniel Friesel-9/+8
2019-08-13Improve _try_fits performance and fairnessDaniel Friesel-0/+3
2019-07-24Improve debuggingDaniel Friesel-1/+5
2019-05-16utils: Add soft_cast_float; float support in config stringsDaniel Friesel-1/+15
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-02 04:37:01 +0000