Feature/to xarray options

[colonesej]

Description

• Allow cli commands to run as python files. Testing and debugging are a bit easier this way. One can simply do

python -m pdb hat/tools/cli.py hat-extract-timeseries config.yml

of with vscode

launch.json

{
    "configurations": [
        {
            "name": "hat-timeseries-extractor",
            "type": "debugpy",
            "request": "launch",
            "program": "${workspaceFolder}/hat/cli.py",
            "console": "integratedTerminal",
            "args": [
                "hat-extract-timeseries",
                "${workspaceFolder}/config.yml",
            ],
            "stopOnEntry": false,
            "justMyCode": false,
        }
    ]
}

• support to_xarray_options in grid options to control xarray dataset creation. Larger datasets would not be loaded lazily going out of memory quite easily for ensemble forecasts/reforecasts.

confuguration looks like

config.yml

station:
  file: "outlets.csv"
  # filters: None
  name: "ObsID"
  coords: 
    x: "LisfloodX"
    y: "LisfloodY"

grid:
  coord_x: "longitude"
  coord_y: "latitude"
  to_xarray_options:
    profile: "mars"
    chunks:
      "longitude": -1
      "latitude": -1
      "time": 1
      "number": 1
  source: 
    file:
      path: "fc.*.grb"
  # source:
  #   mars:
  #     class: ce
  #     date: "20230101"
  #     expver: 1
  #     hdate: ""
  #     levtype: sfc
  #     model: lisflood
  #     number: 1/to/51/by/1
  #     origin: ecmf
  #     param: "240023"
  #     step: "6/12/18/"
  #     stream: efrf
  #     time: "00:00:00"
  #     type: pf
  #     target: "pf_efas5_${date}.grb"

output:
  file: "lisflood_${YMD}.nc"

• Update the expected config structure for extract-timeseries to be all key-value based and the previous one did not support mars type. The down side is having to know the arg names/call signature used in earthkit-data for each source type which can be annoying as they are not consistent, like, .from_source('json', filename) and .from_source('file', path)

Contributor Declaration

By opening this pull request, I affirm the following:

• All authors agree to the Contributor License Agreement.
• The code follows the project's coding standards.
• I have performed self-review and added comments where needed.
• I have added or updated tests to verify that my changes are effective and functional.
• I have run all existing tests and confirmed they pass.

[colonesej]

just FYI, I manually created a pre-release module on the hpc named 0.8pre for Maliko, who needed the chunking option for processing ensemble forecasts data.