py-why/causaltune

Issues

• Add a ranking-based score for ITE problems

  #324 opened 2 months ago by EgorKraevTransferwise
  0

• Speed up calculation of energy score

  #323 opened 2 months ago by EgorKraevTransferwise
  0

• Update to latest versions of DoWhy and EconML

  #322 opened 2 months ago by EgorKraevTransferwise
  0

• Add tests for mac, solve lightgbm install for mac

  #319 opened 2 months ago by EgorKraevTransferwise
  0

• Don't force a time budget at all, and use default n_samples

  #318 opened 2 months ago by ZmeiGorynych
  0

• Correctly handle weights if they are passed to the fit() method of an outcome function

  #316 opened 2 months ago by ZmeiGorynych
  0

• Enable early stopping (and maybe direct categorical feature handling?) in XGBoost etc estimators used for outcome function

  #315 opened 2 months ago by ZmeiGorynych
  0

• Cache part of energy distance

  #287 opened 3 months ago by julianteichgraber
  0

• MSE vs Energy Distance score plot

  #266 opened 3 months ago by julianteichgraber
  0

• Add sklearn estimator selection to config file

  #250 opened 3 months ago by julianteichgraber
  0

• Use EconML bootstrapper SE estimates for custom models

  #243 opened 3 months ago by TDebonoWise
  0

• allow for passing of all/most parameters in both init and fit

  #234 opened 3 months ago by TDebonoWise
  0

• Treat CATE Interpreter as metric

  #220 opened 3 months ago by TDebonoWise
  0

• packaging the library so it can be available via pypi

  #210 opened 3 months ago by yairVanti
  0

• update requirements.txt to latest dowhy/econml releases

  #191 opened 3 months ago by TimoFlesch
  2

• enable confidence intervals for estimators

  #232 opened 3 months ago by TDebonoWise
  0

• move train_test_split to CausalityDataset

  #252 opened 3 months ago by TDebonoWise
  0

• Shap calculation breaking down for |X| > 249

  #240 opened 3 months ago by TDebonoWise
  0

• Evaluation failed error for Quick Start Example

  #297 opened 3 months ago by patriciowoodley
  1

• causatune won't install on python>=3.10

  #305 opened 3 months ago by ZmeiGorynych
  0

• Error when running the ERUPT example notebook

  #304 opened 3 months ago by ZmeiGorynych
  0

• Stop saving copies of the dataset in fitted estimators

  #218 opened 2 years ago by EgorKraevTransferwise
  0

• Recreate sklearn bug: validation check for precomputed gram matrix

  #161 opened 2 years ago by Lekunze
  1

• SimpleParamService doesn't filter estimator list by problem if dataset large and "all" estimators requested

  #188 opened 2 years ago by TimoFlesch
  0

• AttributeError("'NewDummy' object has no attribute 'causal_estimator'"),

  #185 opened 2 years ago by TimoFlesch
  2

• IV Notebook Updates: Comparisons with CATE & SimpleIV, VitD IV example test

  #170 opened 2 years ago by Lekunze
  0

• AutoCausality class should have a method for adding test set scores for each estimator after the fit is done

  #193 opened 2 years ago by EgorKraevTransferwise
  1

• Adapt scoring functions to support multiple treatments

  #195 opened 2 years ago by EgorKraevTransferwise
  2

• Get the Dummy model to support multiple treatments

  #196 opened 2 years ago by EgorKraevTransferwise
  1

• Add an effect() convenience method to AutoCausality class

  #198 opened 2 years ago by EgorKraevTransferwise
  1

• Clean up the dummy models, make sure there's only one and it supports multitreatment

  #214 opened 2 years ago by EgorKraevTransferwise
  1

• Experiment with different propensity functions for the confounded CATE case with synthetic data

  #215 opened 2 years ago by ZmeiGorynych
  0

• Add other scores besides energy_distance to the multitreatment case

  #213 opened 2 years ago by EgorKraevTransferwise
  0

• Make TransformedOutcome model support multivalue treatments

  #212 opened 2 years ago by EgorKraevTransferwise
  0

• Port relevant parts of methodology writeup to workshop template

  #176 opened 2 years ago by EgorKraevTransferwise
  0

• Prepare IV datasets for model validation

  #147 opened 2 years ago by Lekunze
  0

• refactor metrics as dataclass with defined objectives

  #165 opened 2 years ago by Lekunze
  0

• Port abstract and outline into the workshop template

  #175 opened 2 years ago by EgorKraevTransferwise
  0

• Submit PR for using separate validation sets for inner and outer search loops

  #177 opened 2 years ago by EgorKraevTransferwise
  0

• Create IV estimator comparisons with non-linear synthetic dataset

  #181 opened 2 years ago by Lekunze
  0

• Adapt EconML wrapper in DoWhy to handling multiple treatments

  #194 opened 2 years ago by EgorKraevTransferwise
  1

• Front-to-back support for multiple treatments in auto-causality

  #197 opened 2 years ago by EgorKraevTransferwise
  1

• create plots (DAGs) that illustrate CATE/IV problems we study

  #180 opened 2 years ago by TimoFlesch
  0

• write pipeline to run models on synthetic datasets and create plots

  #179 opened 2 years ago by TimoFlesch
  3

• create generator for synthetic datasets with heterogenous treatment effects

  #178 opened 2 years ago by TimoFlesch
  0

• Optimizer doesn't seem to fully respect the time_budget constraint when running with propensity_function='auto'

  #192 opened 2 years ago by EgorKraevTransferwise
  1

• Create synthetic datasets reproducing the structure of our real production ones

  #173 opened 2 years ago by EgorKraevTransferwise
  0

• Independence test-based score for IV models

  #172 opened 2 years ago by EgorKraevTransferwise
  0

• Convert make scorer into independent class

  #157 opened 2 years ago by Lekunze
  0

• Restore correct estimator ATE in Scoring

  #164 opened 2 years ago by Lekunze
  0