pkgdown/extra.css

Skip to contents

lreg_fit() fits a logistic regression model for single predictors. It uses the glm engine for logistic regression and fits the model using the logistic_reg() function from the parsnip package. It also calculates the accuracy, sensitivity, specificity, AUC, and confusion matrix of the model.

Usage

lreg_fit(
  train_data,
  test_data,
  variable = "Disease",
  case,
  cor_threshold = 0.9,
  cv_sets = 4,
  ncores = 1,
  exclude_cols = NULL,
  palette = NULL,
  seed = 123
)

Arguments

train_data

Training data set from make_groups().

test_data

Testing data set from make_groups().

variable

The variable to predict. Default is "Disease".

case

Case to predict.

cor_threshold

Threshold of absolute correlation values. This will be used to remove the minimum number of features so that all their resulting absolute correlations are less than this value.

cv_sets

Number of cross-validation sets. Default is 5.

ncores

Number of cores to use for parallel processing. Default is 4.

exclude_cols

Columns to exclude from the data before the model is tuned. Default is NULL.

palette

Color palette for the ROC curve. Default is NULL.

seed

Seed for reproducibility. Default is 123.

Value

A list with two elements:

  • fit_res: A list with 4 elements:

  • lreg_wf: Workflow object.

  • train_set: Training set.

  • test_set: Testing set.

  • final: Fitted model.

  • metrics: A list with the model metrics:

  • accuracy: Accuracy of the model.

  • sensitivity: Sensitivity of the model.

  • specificity: Specificity of the model.

  • auc: AUC of the model.

  • conf_matrix: Confusion matrix of the model.

  • roc_curve: ROC curve of the model.