summaryrefslogtreecommitdiff
AgeCommit message (Expand)AuthorLines
2024-03-22oooopsmainBirte Kristina Friesel-0/+1
2024-03-22handle mis-specified param filtersBirte Kristina Friesel-1/+5
2024-03-22RMT: add max depth hyper-parameterBirte Kristina Friesel-3/+20
2024-03-21--skip-param-stats and --force-tree are no longer dependentBirte Kristina Friesel-18/+1
2024-03-19dataref export: add preprocessing and RMT method/thresholdBirte Kristina Friesel-6/+29
2024-03-19cli: dataref export: sort by env names rather than values. oops.Birte Kristina Friesel-1/+1
2024-03-12soft_cast_int_or_float: handle values that already are int/floatBirte Kristina Friesel-1/+3
2024-03-12handle floating-point parameter valuesBirte Kristina Friesel-4/+18
2024-03-11CSVfile loader: get ignored and observation (attr) cols from envBirte Kristina Friesel-4/+40
2024-03-11Move Logfile and CSVfile from utils to a new loader classBirte Kristina Friesel-101/+107
2024-03-08build_rmt: handle sklearn.feature_selection.mutual_info_regression failuresBirte Kristina Friesel-3/+9
2024-03-08looks like past-derf actually knew what she was doingBirte Kristina Friesel-0/+1
2024-03-08build_rmt: use DFATOOL_RMT_RELEVANCE_THRESHOLD exclusivelyBirte Kristina Friesel-9/+2
2024-03-08minor warning improvementsBirte Kristina Friesel-9/+9
2024-03-08make FOLFunction inherit from SKLearnRegressionFunctionBirte Kristina Friesel-16/+15
2024-03-08symreg hyper paramBirte Kristina Friesel-11/+15
2024-03-08Add CSV support; right now hard-coded for kaggle acher linux-kernel-sizeBirte Kristina Friesel-3/+31
2024-03-08SKLearnRegressionFunction: export ndarray and pre-proc paramcount to drefBirte Kristina Friesel-45/+73
2024-03-08move fit_parameters check to helper functionBirte Kristina Friesel-60/+26
2024-03-07feature pre-processing via information gainBirte Kristina Friesel-7/+112
2024-03-07RMT: support mutual information-based parameter relevanceBirte Kristina Friesel-1/+28
2024-03-07DFATOOL_RMT_IGNORE_IRRELEVANT_PARAMS=1 → DFATOOL_RMT_RELEVANCE_METHOD=std_b...Birte Kristina Friesel-23/+13
2024-03-07Replace RMT_ENABLED=0 with MODEL=ulsBirte Kristina Friesel-6/+6
2024-03-07dataref export: include environment variablesBirte Kristina Friesel-0/+3
2024-03-07lightgbm: use same logger as dfatoolBirte Kristina Friesel-0/+1
2024-03-06Add LightGBM supportBirte Kristina Friesel-2/+202
2024-03-06boxplot_param: fix --filter-param handlingBirte Kristina Friesel-1/+1
2024-03-06import lightgbmBirte Kristina Friesel-0/+10415
2024-02-26add RMT_SUBMODEL=cartBirte Kristina Friesel-3/+9
2024-02-26Logfile.load: fix errors in error outputBirte Kristina Friesel-3/+2
2024-02-26add --export-csv-unparam supportBirte Kristina Friesel-0/+36
2024-02-23SymbolicRegressionFunction: export hyper-parameters to datarefBirte Kristina Friesel-0/+13
2024-02-22analyze-archive: sorted model outputBirte Kristina Friesel-4/+4
2024-02-22Support Symbolic Regression rather than ULS in RMT leavesBirte Kristina Friesel-16/+29
2024-02-22Replace DFATOOL_RMT_FUNCTION_LEAVES=0 with DFATOOL_RMT_SUBMODEL=staticBirte Kristina Friesel-8/+8
2024-02-21SymbolicRegressor: Add hyper-parameters and complexity score calculationBirte Kristina Friesel-1/+78
2024-02-21SymReg: use MSE and set feature_namesBirte Kristina Friesel-2/+4
2024-02-21black(1)Birte Kristina Friesel-12/+12
2024-02-21make SymbolicRegressionFunction inherit from SKLearnRegressionFunctionBirte Kristina Friesel-113/+39
2024-02-21rename build_dtree to build_rmtBirte Kristina Friesel-10/+10
2024-02-21Rename DFATOOL_FIT_LINEAR_ONLY=1 to DFATOOL_SUBMODEL=folBirte Kristina Friesel-3/+3
2024-02-21Rename DFATOOL_DTREE_* to DFATOOL_RMT_*Birte Kristina Friesel-21/+21
2024-02-21build_dtree: there is no need to explicitly pass param_values and dataBirte Kristina Friesel-8/+2
2024-02-21Replace DFATOOL_FIT_FOL with DFATOOL_MODEL=folBirte Kristina Friesel-41/+47
2024-02-21update documentation and examplesBirte Kristina Friesel-20/+21
2024-02-21move (de)cart, lmt, xgb model generation into separate ModelAttribute functionsBirte Kristina Friesel-84/+99
2024-02-21model: Remove no longer useful build_dtree wrapperBirte Kristina Friesel-39/+24
2024-02-21Always build model in get_fitted; never in constructorBirte Kristina Friesel-40/+23
2024-02-21Move CART/LMT/XGB fitting into the respective modulesBirte Kristina Friesel-217/+270
2024-02-20Re-add (very very basic, for now) Symbolic Regression supportBirte Kristina Friesel-6/+148