CLI ​

HydroPilot exposes four command-line entry points for configuration validation, smoke testing, parameter application, and single-run execution.

Commands ​

Typical workflow ​

For a new configuration, the usual order is:

1. Run hydropilot-validate to confirm the config is structurally correct.
2. Run hydropilot-test to verify that the actual model workflow is wired correctly.
3. Use hydropilot-run or the Python API for controlled evaluation.
4. Use hydropilot-apply when you want a materialized project copy with parameters written in.

hydropilot-validate ​

hydropilot-validate <config.yaml>

Use this command to check a config without running the model. It parses YAML, expands template configs when needed, validates the resulting general config, and prints diagnostics line by line.

Important points:

• exit code 0 means validation passed, even if warnings are present
• exit code 1 means at least one error-level diagnostic
• the command is read-only and does not modify project files

hydropilot-test ​

hydropilot-test <config.yaml>

This is the full smoke-test path. HydroPilot loads the config, builds a default parameter vector from parameter bounds, creates a temporary project copy, runs the model once, extracts outputs, and evaluates objectives and diagnostics.

Use it when you need to verify that:

• project paths are correct
• the model command actually runs
• output files are readable
• evaluation metrics are wired correctly

The smoke test forces parallel = 1 and keepInstances = true so the run is easier to inspect.

hydropilot-apply ​

hydropilot-apply <apply.yaml>

This command creates a fresh project copy and writes parameters into it.

The apply file normally includes:

config: path/to/config.yaml
mode: design
values:
  CN2: 72.5
  ALPHA_BF: 0.3
  GW_DELAY: 120
outDir: ./applied_project

Use mode: design when values live in design space and must be transformed before writing. Use mode: physical when you already have resolved physical parameters.

hydropilot-run ​

hydropilot-run <run.yaml>

This is the single-run entry point. It loads the referenced config, evaluates one parameter vector, and prints a summary of objectives, constraints, diagnostics, and output locations.

It is intentionally narrow in scope:

• good for one-off evaluations
• not intended for optimization campaigns
• not intended for experiment management

For repeated runs or algorithm-driven workflows, move to SimModel or UQPyLAdapter.

What HydroPilot CLI is good at ​

• early config diagnostics
• smoke testing a real project setup
• parameter application for inspection
• single-run execution without writing custom Python

Command comparison ​

Read next ​

• Examples
• Python API