Reference

All functions

apply_basic_recipe()

Apply basic recipe to the dataframe that includes a text column. The basic recipe includes tokenization (using bigrams), removing stop words, filtering stop words by max tokens = 1,000, and normalization of document length using TF-IDF.

build_pipeline()

Build a pipeline from creating tuning parameters, search spaces, workflows, 10-fold cross-validation samples to finding the best model from lasso, random forest, XGBoost to fitting the best model from each algorithm to the data

cal_class_fit()

Evaluate a classification model output

create_cv_folds()

Create 10-fold cross-validation samples

create_search_spaces()

Create search spaces for the algorithms based on the hyperparameters

create_tunes()

Create tuning parameters for algorithms (i.e., lasso, random forest, and XGBoost).

create_workflows()

Create workflows for the algorithms based on the hyperparameters

find_best_model()

Find the best version of each algorithm based on the hyperparameters and 10-fold cross-validation.

fit_best_model()

Fit the best model from each algorithm to the data.

split_using_srs()

Creating training and testing data based on stratified random sampling (SRS) and preprocessing steps

topn_vip()

Visualize the importance of top 20 features

viz_class_fit()

Visualize a classification model output