Commands:
  account            Create an account for a tenant in the FlexMeasures...
  account-role       Create an account role.
  annotation         Add annotation to accounts, assets and/or sensors.
  asset              Add an asset.
  asset-type         Add an asset type.
  beliefs            Add sensor data from a CSV or Excel file.
  forecasts          Generate forecasts for a target sensor.
  holidays           Add holiday annotations to accounts and/or assets.
  initial-structure  Initialize useful structural data.
  report             Create a new report using the Reporter class and...
  schedule           (Deprecated) Create a new schedule for a given power...
  sensor             Add a sensor.
  source
  sources            Create data sources for the data generators found...
  toy-account        Create a toy account, for tutorials and trying things.
  user               Create a FlexMeasures user.

Usage: flexmeasures add forecasts [OPTIONS]

  Generate forecasts for a target sensor.

  Example
    flexmeasures add forecasts --sensor 2092 --regressors 2093
      --start-date 2025-01-01T00:00:00+01:00 --to-date 2025-10-15T00:00:00+01:00

  Workflow
    - Training window: defaults from --start-date until the CLI execution time.
    - Prediction window: defaults from CLI execution time until --to-date.
    - max-forecast-horizon: defaults to the length of the prediction window.
    - Forecasts are computed immediately; use --as-job to enqueue them.
    - Sensor 2093 is used as a regressor in this example.

  Notes:
  - Use --from-date to explicitly set when the forecasts will start.
  - Use --train-period to set the training window, which will grow each cycle
      until the specified --to-date is reached.
  - Use --predict-period to set the prediction window. It rolls forward by the
      forecast period each cycle, similar to the training window, but its size
      does not grow.

Options:
  --sensor TEXT                   Create forecasts for this sensor. Follow up
                                  with the sensor's ID. This argument can be
                                  given multiple times.  [required]
  --regressors TEXT               Comma-separated list of sensor IDs to be
                                  used as regressors. This is the full set of
                                  regressors and can include both past
                                  (realizations) and future (forecasts).
                                  Default is 'autoregressive', which uses the
                                  target sensor's own history.
  --future-regressors TEXT        Comma-separated list of sensor IDs to be
                                  treated as future regressors. Future
                                  regressors are assumed to have forecasts
                                  only.
  --past-regressors TEXT          Comma-separated list of sensor IDs to be
                                  treated as past regressors. Past regressors
                                  are assumed to have realizations only.
  --train-start, --start-date TEXT
                                  Start date for running the pipeline, i.e.
                                  when training begins (format: YYYY-MM-
                                  DDTHH:MM:SS±HH:MM).  [required]
  --to-date, --end-date TEXT      End date for running the pipeline (YYYY-MM-
                                  DDTHH:MM:SS+HH:MM)  [required]
  --train-period INTEGER RANGE    Duration of the initial training period in
                                  days (minimum of 2). After each forecast
                                  period, each next cycle will increase the
                                  training period by the forecast period. If
                                  not set, derives a training period from
                                  --start-predict-date instead. If that is
                                  also not set, defaults to 2 days.  [x>=2]
  --predict-period TEXT           Number of days to predict into the future in
                                  each cycle. After each cycle, the prediction
                                  window will move forward by this number of
                                  days.
  --from-date TEXT                Start date for predictions (YYYY-MM-
                                  DDTHH:MM:SS+HH:MM)
  --max-forecast-horizon TEXT     Maximum forecast horizon in hours
  --forecast-frequency INTEGER    Forecast frequency in hours, i.e. how often
                                  to recompute forecasts.
  --model-save-dir TEXT           Directory to save the trained model
  --output-path TEXT              Directory to save prediction outputs
  --probabilistic                 Enable probabilistic predictions
  --sensor-to-save TEXT           Sensor ID to save forecasts into a specific
                                  sensor. By default, forecasts are saved to
                                  the target sensor.
  --as-job                        Whether to queue a forecasting job instead
                                  of computing directly. To process the job,
                                  run a worker (on any computer, but
                                  configured to the same databases) to process
                                  the 'forecasting' queue. Defaults to False.
  --help                          Show this message and exit.