v0.1.0

There are a number of changes with this release:

• Lots of changes with PR #44:
  • Added new functions:
    • get_fcast_definition, which stores and returns forecast schedule definitions and supports adjust_forecast_datetimes
    • adjust_forecast_datetimes which takes in a target available date and lead time/run length and returns an init date (and lead time/run length) that represents the latest model outputs available for the selected model (closes #31)
  • Removed some functionality from model_input_formatter - it no longer rounds the init_date down to the last valid cycle time for the model. Users should use adjust_forecast_datetimes instead if they are not sure about the correct init date (Closes #33, #12, #11)
  • Created new model names for IFS Ensemble ('ifs_ens') and AIFS Ensemble ('aifs_ens') to differentiate them from the Single versions (Closes #39)
  • Closed #27 with 9efed98
  • Added a section to the readme on handling dates/times
• Fixed #38 so that requests for irradiance in GFS at f000 with be corrected and hefty will issue a warning
• Improved handling of Herbie/FastHerbie errors, closing #1
• Fixed an issue with newer versions of eccodes that broke get_wind_forecast in PR #42, closing issue #41 (and adding more models to the wind example notebook)
• Added 2m air temperature back for GFS in get_wind_forecast

Breaking changes:

model_input_formatter no longer rounds the init_date down to the last valid cycle time for the model. Users should use adjust_forecast_datetimes instead if they are not sure about the correct init date