Python CLI¶

The Python package installs a CLI entry point:

regimeflow-backtest

It is implemented in python/regimeflow/cli.py and is intended for Python-facing backtest runs that export reports and summaries.

What The CLI Is For¶

Use the CLI when you want:

a reproducible command for a Python-run backtest
report export without writing a wrapper script
a quick bridge from YAML config to JSON/CSV outputs
Python strategy loading through module:Class

Command Form¶

regimeflow-backtest --config <path> --strategy <strategy>

Equivalent module form:

.venv/bin/python -m regimeflow.cli --config <path> --strategy <strategy>

Arguments¶

Argument	Meaning
`--config`	YAML backtest config path
`--strategy`	Built-in/registered strategy name, or `module:Class` for a Python strategy
`--strategy-params`	Optional JSON file merged into `strategy_params`
`--output-json`	Write serialized report JSON
`--output-csv`	Write serialized report CSV
`--output-equity`	Write equity curve CSV
`--output-trades`	Write trades CSV
`--print-summary`	Print performance summary JSON to stdout

Built-In Strategy Example¶

regimeflow-backtest \
  --config examples/backtest_basic/config.yaml \
  --strategy moving_average_cross \
  --output-json report.json \
  --output-equity equity.csv \
  --output-trades trades.csv \
  --print-summary

Python Strategy Example¶

regimeflow-backtest \
  --config examples/backtest_basic/config.yaml \
  --strategy my_strategies:MyStrategy

The class must:

be importable from the given module
subclass regimeflow.Strategy

Strategy Parameters¶

If you want to inject strategy_params from a JSON file:

regimeflow-backtest \
  --config examples/backtest_basic/config.yaml \
  --strategy my_strategies:MyStrategy \
  --strategy-params params.json

The JSON file must contain an object.

What The CLI Returns¶

On success:

exit code 0
optional files written to the requested paths
optional summary printed to stdout

On failure:

exit code 1
a human-readable error message to stderr

When To Use The CLI Vs Python API¶

Use the CLI when:

you want a single shell command for a run
your configuration is already YAML-driven
your output is report/export oriented

Use the Python API directly when:

you are iterating inside notebooks or scripts
you want multi-step programmatic control
you need direct access to results, dataframes, NumPy arrays, or dashboards

python/overview.md
python/workflow.md
tutorials/python-usage.md