Configuration recipe catalog

Copy a YAML template from the repository config/ tree, edit the #%%==== path blocks, then run the matching habit command. Relative paths in YAML resolve from the config file directory, not the shell working directory.

Note

The lightweight Windows ZIP ships config/ next to the installers. Source checkouts keep the same layout under the repository root.

By workflow

Workflow	Template directory	Typical command
Preprocessing	`config/preprocessing/`	`habit preprocess -c …`
DICOM sort	`config/dicom_sort/`	`habit sort-dicom -c …`
Habitat segmentation	`config/habitat/`	`habit get-habitat -c …`
Habitat feature extraction	`config/feature_extraction/`	`habit extract -c …`
Traditional radiomics	`config/radiomics/`	`habit radiomics -c …`
Machine learning	`config/machine_learning/`	`habit model -c …` / `habit cv -c …`
Model comparison	`config/model_comparison/`	`habit compare -c …`
ICC / test-retest	`config/auxiliary/`	`habit icc -c …` / `habit retest -c …`

Starter demos

Use these when you want a short path through the demo data:

Goal	Suggested YAML
Resample-only preprocess	`config/preprocessing/config_preprocessing_demo.yaml`
Habitat two-step train	`config/habitat/config_habitat_two_step.yaml`
Extract habitat features	`config/feature_extraction/config_extract_features_demo.yaml`
Traditional ROI radiomics	`config/radiomics/config_traditional_radiomics.yaml`
Train a radiomics ML model	`config/machine_learning/config_machine_learning_radiomics.yaml`
K-fold CV demo	`config/machine_learning/config_machine_learning_kfold_demo.yaml`
Compare models	`config/model_comparison/config_model_comparison_demo.yaml`
ICC demo	`config/auxiliary/config_icc_demo.yaml`

Field reference pages