params.yaml

base:
  random_state: 42
  log_level: INFO

data_load:
  dataset_zip: health-dataset.zip
  dataset_folder: data/raw
  dataset_csv: data/raw/diabetes_data.csv

featurize:
  feature_folder: data/processed
  features_path: data/processed/featured_data.csv
  drop_features:
  - Veggies
  target_column: Diabetes

data_split:
  test_size: 0.3
  trainset_path: data/processed/train.csv
  testset_path: data/processed/test.csv

train:
  cv: 3
  estimator_name: xgb
  estimators:
    logreg:
      param_grid:
        C: [0.1]
        max_iter: [200]
        solver: [lbfgs]
        multi_class: [multinomial]
    knn:
      param_grid:
        n_neighbors: [9]
    xgb:
      param_grid:
        min_child_weight: [1]
        gamma: [1, 5]
        subsample: [1.0]
        colsample_bytree: [0.6, 1.0]
        max_depth: [3]
    dectree:
      param_grid:
        criterion: [gini, entropy]
        max_depth: [8]
    randforest:
      param_grid:
        n_estimators: [100]
        max_depth: [11]
        min_samples_split: [15]
  model_folder: model
  model_path: model/model.joblib
  scaler_path: model/scaler.joblib

evaluate:
  report_folder: report
  accuracy_report_path: report/report.json
  confusion_matrix_path: report/confusion_matrix.png
  translated_report_path: report/confusion_matrix_data.csv
  target_names:
  - no_diabetes
  - diabetes