Getting Started¶

Installation¶

pip install simnibs-analyze

Pulls in simnibs-reader, which handles all NIfTI reading and ROI extraction.

You need SimNIBS outputs already computed: a head model per subject (m2m_<subID>/, from charm) and a simulation or optimization folder.

The two commands¶

Both are driven entirely by a YAML file passed with --config. The schemas are different and not interchangeable.

`simnibs-analyze` — features and statistics¶

simnibs-analyze --config config-analyze.yaml

Per (subject × condition × mode) it extracts the target ROI, post-processes it, computes intra-ROI stats, extra-ROI stats and their ratio, then appends a row to all_features_space-<space>.csv. It finishes with the inter/intra-subject analysis and clustering, writing to paths.out_root.

Feature extraction is the slow part. Once the CSV exists, --skip-features re-runs only the statistics; --skip-analysis stops after extraction.

→ Configuration reference · Outputs

`simnibs-viz` — figures and cohort montages¶

simnibs-viz --config config-viz.yaml

Renders one figure per subject for each figure block, then composes cohort montages for the blocks flagged cohort: true, on a single shared e-field scale. Output goes to paths.out_root, one directory per {subject}_{simulation}, with montages in _cohort/.

→ Configuration reference · Outputs

Checking a config before running¶

python -m simnibs_analyze._config_schema_analyze --config my-config.yaml

Prints a one-line summary, or the validation error and exit code 1.

Development¶

git clone https://github.com/ICM-Frontlab-CDPR/simnibs-analyze.git
cd simnibs-analyze
pip install -e ".[dev]"
pytest