Evaluation

Use bakaano.neuralnet.simulate.evaluate_streamflow_model_interactively to compare a trained model against one station at a time.

Minimal working example

from bakaano.neuralnet.simulate import evaluate_streamflow_model_interactively

evaluate_streamflow_model_interactively(
    working_dir="/path/to/working_dir",
    study_area="/path/to/basin.shp",
    model_path="/path/to/working_dir/models/bakaano_model.keras",
    val_start="2013-01-01",
    val_end="2020-12-31",
    grdc_netcdf="/path/to/GRDC.nc",
    routing_method="mfd",
)

Evaluate with station CSVs

evaluate_streamflow_model_interactively(
    working_dir="/path/to/working_dir",
    study_area="/path/to/basin.shp",
    model_path="/path/to/working_dir/models/bakaano_model.keras",
    val_start="2013-01-01",
    val_end="2020-12-31",
    grdc_netcdf=None,
    csv_dir="/path/to/observed_csvs",
    lookup_csv="/path/to/station_lookup.csv",
)

How it works

• Loads one trained model.

• Builds predictors for the requested evaluation period.

• Prompts you to select a station.

• Plots observed versus predicted streamflow.

Common failure modes

• The model file does not exist.

• The requested validation period is not covered by routed runoff.

• CSV observations are incomplete or use different station IDs than the lookup table.

• The trained model does not have a sidecar config file and fallback scaling options were not set to match training.