Package index • patroklos

All functions

Ass2d: An R6 class to assess a model with continuous output with one metric versus another.

AssScalar: An R6 class to assess a model with scalar metrics

Data: An R6 class for the data

Model: An R6 class for a model

discretize_tbl_cols(): Discretize columns of a tibble and append the resulting columns

ensure_patients_match(): Ensure patients in expression and pheno tibbles match during preprocessing

error_rate(): Error rate of a binary classifier

fitter_prototype(): Function interface of a fitter

greedy_nestor(): Nest an existing, tuned model together with more features into another model and tune the latter

imputer_prototype(): Function interface of an imputer

long_nestor(): Nested cross-validation for second-stage validated predictions

mean_impute(): Impute missing values in a matrix by column means

multitune(): Tune multiple hyperparameters with a single call to a fitter

multitune_output_prototype(): Function interface for the return value of multitune()

neg_binomial_log_likelihood(): Negative binomial log-likelihood

neg_prec_with_prev_greater(): Minimal negative precision for thresholds with a minimal prevalence

neg_roc_auc(): Negative ROC AUC

nested_fit(): Construct a nested_fit S3 object

non_zero_coefs(): Get features with non-zero coefficients

predict(<nested_fit>): Predict method for nested_fit objects

predict(<ptk_ranger>): Wrap ranger::predict.ranger() into a patroklos-compliant predict function

predict(<ptk_zerosum>): Wrap zeroSum::predict.zeroSum() into a patroklos-compliant predict function

predict_method_prototype(): Function interface of the S3 method predict() for a fit_obj

prepend_to_directory(): To every Model in a list of Models, prepend a fixed directory to the directory attribute

prepend_to_filename(): Prepend a prefix to the file attribute of a list of AssScalar objects

projection_on_feature(): Project on a feature

ptk_ranger(): Wrap ranger::ranger() into a patroklos-compliant fitter

ptk_zerosum(): Wrap zeroSum::zeroSum() into a patroklos-compliant fit function

training_camp(): Wrap Model$fit() to fit, validate and store multiple models

val_error_fun_prototype(): Function interface of a validation error function

val_vs_test(): Compare validation and test error

write_data_info(): Write data info to a JSON file