Weather data is now loaded from two sources. If an avalanche bulletin exists for a date,
the MountainWeather attributes on the bulletin is used. If no bulletin exists, the data is
fetched from the weather API.
ForecastDataset.label() now have a parameter called with_varsom that defaults to
True. If set to false, avalanche warning attributes are removed from the input dataset.
Downloads are parallelized. Right now it uses requests-futures, but this may change in the
future, when a similar change is made for varsomdata, but without external dependencies.
The effectiveness of this seems to vary depending on system
I have not completed the filtering/selection of specific columns for the outdata yet.
That's next up.
Weather data is now loaded from two sources. If an avalanche bulletin exists for a date, the MountainWeather attributes on the bulletin is used. If no bulletin exists, the data is fetched from the weather API.
ForecastDataset.label()
now have a parameter calledwith_varsom
that defaults toTrue
. If set to false, avalanche warning attributes are removed from the input dataset.Downloads are parallelized. Right now it uses
requests-futures
, but this may change in the future, when a similar change is made forvarsomdata
, but without external dependencies. The effectiveness of this seems to vary depending on systemI have not completed the filtering/selection of specific columns for the outdata yet. That's next up.