Latest Forecast Downloader (2DMI)

Standalone tooling for downloading GHM timeseries data.

The repository currently includes a script for downloading the latest forecast for a list of PFAF IDs, and is structured so additional scripts can be added over time.

What download_latest_forecast.py does

1. Calls /subProjectIds for the root project.
2. Picks the subproject with the largest timestamp (latest forecast).
3. Calls bulk /ghm-timeseries/tsIds for requested PFAF IDs.
4. Saves each series to CSV in the output folder.

Repository Files

• download_latest_forecast.py: standalone CLI script for latest forecast download
• dhi-ghm-downloader/src/: reusable package modules (api, forecast, pfaf, dataframe helpers)
• pixi.toml: Pixi environment and run tasks
• requirements.txt: optional pip dependency file

Setup

pixi install

Run Latest Forecast Script

--project-id is required.

At least one PFAF input is required:

• --pfaf-ids
• --pfaf-file

If neither is provided, the script exits with an input error:

Input error: Provide PFAF IDs with --pfaf-ids and/or --pfaf-file

Use space-separated PFAF IDs:

pixi run python download_latest_forecast.py --api-key <YOUR_API_KEY> --project-id <ROOT_PROJECT_ID> --pfaf-ids 136320200000 219730300000

Use list-style PFAF IDs:

pixi run python download_latest_forecast.py --api-key <YOUR_API_KEY> --project-id <ROOT_PROJECT_ID> --pfaf-ids "[136320200000,219730300000]"

Use PFAF IDs from file:

pixi run python download_latest_forecast.py --api-key <YOUR_API_KEY> --project-id <ROOT_PROJECT_ID> --pfaf-file pfaf_ids_example.txt

Use predefined Pixi task:

pixi run download_latest --api-key <YOUR_API_KEY> --project-id <ROOT_PROJECT_ID> --pfaf-file pfaf_ids_example.txt

Arguments

Required arguments:

• --api-key
• --project-id
• At least one of: --pfaf-ids or --pfaf-file

Optional arguments:

• --pfaf-ids
• --pfaf-file
• --base-url (default: https://wrd-mike-cloud-performance-statistics.azurewebsites.net/api)
• --environment (default: prod)
• --timeout (default: 60)
• --outdir (default: output_latest_forecast)

Adding More Scripts in the Future

To keep Pixi usage consistent as more Python scripts are added:

1. Add the script at repository root (or preferred scripts folder).
2. Register a dedicated task in pixi.toml, for example:

[tasks]
download_latest = "python download_latest_forecast.py"
download_historical = "python download_historical_forecast.py"

3. Run with the same pattern:

pixi run <task_name> --<script-args>

Keep the generic python = "python" task for ad hoc script execution.