SchlossLab/mikropml

Issues

• run_ml does not support "training_frac = 1.0" to get best hyperparameter value

  #345 opened 4 months ago by joannacolovas
  0

• Preprocessing in run_ml?

  #343 opened 5 months ago by pschloss
  1

• Precision should be 1 when TP=1 and FP=0

  #338 opened 2 years ago by kelly-sovacool
  0

• Factor levels in outcome column are not preserved

  #339 opened 2 years ago by kelly-sovacool
  0

• Calculate the area under the balanced precision-recall curve (AUBPRC)

  #330 opened 2 years ago by kelly-sovacool
  0

• roxygen2 docs: switch `@noRd` to `@keywords internal` for all internal functions

  #328 opened 2 years ago by kelly-sovacool
  0

• New vignette for visualizing performance and feature importance

  #334 opened 2 years ago by kelly-sovacool
  0

• Impute missing values after the train/test split

  #301 opened 2 years ago by kelly-sovacool
  0

• Report confidence interval for permutation feature importance

  #324 opened 2 years ago by kelly-sovacool
  0

• Switch from {future.apply} to {furrr} where possible

  #327 opened 2 years ago by kelly-sovacool
  1

• Create function to bootstrap the test set to get a confidence interval for performance

  #325 opened 2 years ago by kelly-sovacool
  0

• use which run_ml() output to plot ROC curve?

  #296 opened 3 years ago by zhq90
  2

• Add ROC curve example code to docs

  #262 opened 2 years ago by kelly-sovacool
  3

• NaN as character causes some problems

  #313 opened 2 years ago by sskoldas
  3

• Extracting weights from 100 iterations and from random random forset model

  #319 opened 2 years ago by Hesham999666
  1

• Use hyperparams from `rf` for `parRF`

  #305 opened 2 years ago by kelly-sovacool
  0

• meta-analysis using mikropml?

  #308 opened 2 years ago by marwa38
  0

• Training Webinar

  #241 opened 2 years ago by SWi1
  2

• Warning message about '%dorng%' vs '%dopar%' when run in parallel

  #283 opened 2 years ago by dansmith01
  2

• Providing case weights

  #271 opened 2 years ago by funnell
  4

• Update from model formula to x,y notation in train function

  #242 opened 2 years ago by BTopcuoglu
  0

• Refactor `select_apply()`

  #302 opened 2 years ago by kelly-sovacool
  0

• Need function to compare models

  #268 opened 3 years ago by pschloss
  5

• classification of new (and outcome unknown) samples

  #294 opened 3 years ago by fairquestion
  1

• Report p-values for permutation importance

  #287 opened 3 years ago by kelly-sovacool
  0

• group_correlated_features() sometimes appends extra columns in R 3.6.1

  #264 opened 3 years ago by kelly-sovacool
  0

• grouped features are not always in the same order

  #269 opened 3 years ago by ecmaggioncalda
  0

• Better explain the groups param & batch effects in the intro vignette

  #234 opened 4 years ago by kelly-sovacool
  0

• Note --max-ppsize for large datasets in preprocess vignette

  #233 opened 4 years ago by kelly-sovacool
  0

• find_feature_importance fails for multiclass data

  #274 opened 3 years ago by alexmsalmeida
  6

• Conversion of numeric outcomes to character error

  #272 opened 3 years ago by ecmaggioncalda
  1

• update get_corr_feats() to have correlation method option

  #265 opened 4 years ago by courtneyarmour
  1

• Correlation threshold parameter

  #263 opened 4 years ago by kelly-sovacool
  1

• Implement progress bar for long-running functions

  #249 opened 4 years ago by kelly-sovacool
  3

• Test multi-outcome datasets

  #252 opened 4 years ago by kelly-sovacool
  2

• Use the phrase "model training" rather than "ML" where appropriate

  #253 opened 4 years ago by kelly-sovacool
  0

• Report live updates of metrics during training

  #250 opened 4 years ago by kelly-sovacool
  1

• Refactor run_ml() API for future expandability

  #251 opened 4 years ago by kelly-sovacool
  1

• Class strings cannot contain spaces in the outcome columns [JOSS review]

  #244 opened 4 years ago by JonnyTran
  2

• decision tree error

  #243 opened 4 years ago by ebechtold
  4

• Show how to do feature importance after run_ml() if you skipped it initially

  #235 opened 4 years ago by kelly-sovacool
  1

• Support custom groups of features for permutation importance

  #245 opened 4 years ago by kelly-sovacool
  0

• Intro vignette proof read

  #231 opened 4 years ago by pschloss
  0

• Error: "long vectors not supported" with very large datasets

  #218 opened 4 years ago by zenalapp
  1

• Preprocessing Data Vignette text correction

  #228 opened 4 years ago by courtneyarmour
  0

• Continuous data normalization methods

  #229 opened 4 years ago by lucas-bishop
  1

• Edit hyperparameter tuning vignette

  #230 opened 4 years ago by NLesniak
  3

• Feature importances are wrong when run_ml() is given a tibble instead of a dataframe

  #225 opened 4 years ago by kelly-sovacool
  0

• make otu_mini slightly larger

  #214 opened 4 years ago by zenalapp
  1

• profile preprocess_data to try to make it faster

  #215 opened 4 years ago by zenalapp
  1