CLI Reference

PhenoCluster provides a command-line interface for running the pipeline, generating configuration files, and validating configurations.

phenocluster

PhenoCluster - Clinical Phenotype Discovery Pipeline

Usage

phenocluster [OPTIONS] COMMAND [ARGS]...

create-config

Generate a configuration YAML file from a profile template.

Profiles set sensible defaults for common use-cases. Data-specific parameters (column names, survival targets) are left as placeholders that you fill in.

Example:

phenocluster create-config -p complete -o config.yaml phenocluster create-config -p quick -o quick_config.yaml

Usage

phenocluster create-config [OPTIONS]

Options

-o, --output <output>

Output YAML path

Default:

'config.yaml'

-p, --profile <profile>

Profile template to use

Default:

<ProfileName.complete: 'complete'>

Options:

descriptive | complete | quick

dashboard

Launch an interactive Streamlit dashboard over saved pipeline outputs.

Requires the optional ‘dashboard’ extras: pip install ‘phenocluster[dashboard]’

Example: phenocluster dashboard ./results/

Usage

phenocluster dashboard [OPTIONS] RESULTS_DIR

Options

--port <port>

Local port to bind the dashboard.

Default:

8501

--host <host>

Host interface to bind.

Default:

'127.0.0.1'

--headless, --browser

Run in headless mode (do not auto-open a browser).

Default:

True

Arguments

RESULTS_DIR

Required argument

Path to a previously saved PhenoCluster results directory.

list-profiles

List available configuration profile templates.

Usage

phenocluster list-profiles [OPTIONS]

run

Run the phenotype discovery pipeline.

All parameters are controlled via the configuration YAML file. Use ‘create-config’ to generate a config from a profile.

Example:

phenocluster run -d data.csv -c config.yaml phenocluster run -d data.csv -c config.yaml –force-rerun

Usage

phenocluster run [OPTIONS]

Options

-d, --data <data>

Required Path to input CSV file

-c, --config <config>

Required Path to configuration YAML file

--force-rerun

Ignore cached artifacts and re-run all steps

Default:

False

-v, --verbose

Enable DEBUG logging and full tracebacks on error.

Default:

False

-q, --quiet

Suppress the banner and non-essential output.

Default:

False

--html-report, --no-html-report

Generate the static HTML analysis report. Overrides config.generate_html_report.

show-profile

Print the resolved YAML for a profile with syntax highlighting.

Usage

phenocluster show-profile [OPTIONS] PROFILE:{descriptive|complete|quick}

Arguments

PROFILE

Required argument

Profile name

validate-config

Validate a configuration YAML file.

Checks YAML structure, required sections, value ranges, and internal consistency. When –data is supplied, also cross-references every column name in the config against the actual CSV header.

Examples:

phenocluster validate-config -c config.yaml phenocluster validate-config -c config.yaml -d data.csv

Usage

phenocluster validate-config [OPTIONS]

Options

-c, --config <config>

Required Path to configuration YAML file

-d, --data <data>

Path to CSV file - cross-checks column names against actual data

version

Show version information.

Usage

phenocluster version [OPTIONS]