# Outputs

Two main kinds of output can be produced by the `Physprep` workflow: raw physiolocal data
organized in BIDS format (according to the BIDS specification v1.10.0), and physiological
derivatives.

## Raw physiological data

If the `skip_convert` flag is not specified when running the workflow, all of the output
will be save at the participant level under `sub-<sub_id>/ses-<ses_id>/<func/eeg>`. If the
physiological data were acquired concurrently with fmri data, the output will be save
under the `func` subfolder. If they were acquired concurrently with EEG data, the output
will be save under the `eeg` subfolder.

### Data and metadata files

For each session (or run) the following files will be generated:

- `[matches]_physio.tsv.gz` : raw segmented physiological timeseries.
- `[matches]_physio.json` : metadata file containing the tsv columns names, start time, and
signal sampling frequency information.

## Physiological derivatives

If `derivatives_dir` is specified when running the workflow, all the derivatives will be
save at the participant level in that directory. If not specified, the derivatives will
be save at the participant level under `<indir_bids>/derivatives/physprep`.

### Data files

For each session (or run) the following files will be generated:

- `[matches]_desc-preproc_physio.tsv.gz` : preprocessed timeseries.
- `[matches]_desc-quality.json` : quality assessment on the preprocessed timeseries.
- `[matches]_desc-physio_events.tsv` : extracted features.

### Code files

At the root of the output directory, the following files will be generated under `code/`:

- `log.txt`: log file.
- `preprocessing_strategy.json`: preprocessing steps applied to each physiological modality.
- `emissions.csv`: file generated by `codecarbon`. Will be generated only if `--track_carbon` is specified (see {ref}`carbon-tracker`).