CreateLogisticRegressionModel — Analysis Tool Box

Introduction

When the outcome is binary — churn or not, default or not, approve or reject — the useful answer usually isn't just a label but a calibrated probability you can act on. CreateLogisticRegressionModel fits a scikit-learn logistic regression for exactly that: it cleans the data with complete-case analysis, optionally scales features, splits train/test, and generates a confusion-matrix heatmap plus a training-vs-test error-rate comparison, giving you a fast, well-calibrated baseline classifier with a visible read on overfitting.

Teaching Note

Logistic regression is essential for:

Predicting the likelihood of customer conversion or subscription renewal
Assessing the probability of default in credit risk analysis
Identifying the drivers behind binary choices in consumer behavior
Classifying medical or technical reports into specific categories (e.g., critical/standard)
Forecasting the outcome of competitive bids or project approvals
Evaluating the impact of different features on a categorical classification
Building efficient baseline classifiers for machine learning pipelines

Parameters

Parameter	Default	Description
`dataframe`required	—	The input pandas DataFrame containing the training and testing data.
`outcome_variable`required	—	The name of the categorical target column to be predicted.
`list_of_predictor_variables`required	—	A list of column names used as features for the classification.
`scale_predictor_variables`	`False`	If True, scales the features using StandardScaler prior to training. Defaults to False.
`print_peak_to_peak_range_of_each_predictor`	`False`	If True, prints the statistical range of each predictor column to monitor data dispersion. Defaults to False.
`test_size`	`0.2`	The proportion of the dataset used for testing. Defaults to 0.2.
`show_classification_plot`	`True`	Whether to display a heatmap of the confusion matrix labeled with the overall accuracy score. Defaults to True.
`lambda_for_regularization`	`0.001`	The regularization parameter. Internally mapped to C = 1 - lambda. Lower values increase regularization strength. Defaults to 0.001.
`max_iterations`	`1000`	The maximum number of iterations allowed for the solver to converge. Defaults to 1000.
`random_seed`	`412`	Controls the randomness of the data split and solver initialization. Defaults to 412.
`print_model_training_performance`	`False`	If True, prints training error rate and test error rate after fitting. Defaults to False.
`plot_training_and_test_performance`	`True`	Whether to render a bar chart comparing training error rate and test error rate. Defaults to True.
`training_bar_color`	`'#3a86ff'`	Bar color for the training error rate bar. Defaults to '#3a86ff'.
`test_bar_color`	`'#b0170c'`	Bar color for the test error rate bar. Defaults to '#b0170c'.
`figure_size_for_performance_comparison_plot`	`(7, 5)`	Dimensions (width, height) for the comparison chart. Defaults to (7, 5).
`title_for_performance_comparison_plot`	`None`	Title text for the comparison chart. Defaults to a sensible built-in title when None.
`subtitle_for_performance_comparison_plot`	`None`	Subtitle text for the comparison chart. Defaults to a sensible built-in subtitle when None.
`caption_for_performance_comparison_plot`	`None`	Optional caption text for the comparison chart. Defaults to None.
`title_y_indent_for_performance_comparison_plot`	`1.1`	Vertical position of the title on the comparison chart. Defaults to 1.1.
`subtitle_y_indent_for_performance_comparison_plot`	`1.05`	Vertical position of the subtitle on the comparison chart. Defaults to 1.05.
`caption_y_indent_for_performance_comparison_plot`	`-0.15`	Vertical position of the caption on the comparison chart. Defaults to -0.15.
`x_indent_for_performance_comparison_plot`	`-0.115`	Horizontal starting position for text on the comparison chart. Defaults to -0.115.

Returns

If scale_predictor_variables=False, returns the fitted sklearn.linear_model.LogisticRegression. If True, returns a dictionary containing both the 'model' and the 'scaler' object so new observations can be scaled consistently before prediction.

Example

python

from analysistoolbox.predictive_analytics import CreateLogisticRegressionModel

# Basic model to predict customer churn
model = CreateLogisticRegressionModel(
    df,
    outcome_variable='is_churn',
    list_of_predictor_variables=['tenure', 'monthly_spend'],
)