You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now, evalcast::get_predictions() downloads data from covidcast, then applies a single forecaster to the data.
When doing forecast development, it is common to try out many different forecasters on the same set of data. get_predictions() is too slow for this at the moment as it has to download the same set of data for each forecaster, and downloading of the data takes up a significant chunk of the runtime. (This is because we have to download several months worth of data to make corrections.)
It would be great if we only have to download the data once for multiple forecasters. I think there are (at least) two ways to do this:
Modify get_predictions() so that it can take in a list of forecasters.
Break get_predictions() into two functions: one that gets the relevant data, and one that gets predictions with the data from the first function as input.
The text was updated successfully, but these errors were encountered:
Right now,
evalcast::get_predictions()
downloads data fromcovidcast
, then applies a single forecaster to the data.When doing forecast development, it is common to try out many different forecasters on the same set of data.
get_predictions()
is too slow for this at the moment as it has to download the same set of data for each forecaster, and downloading of the data takes up a significant chunk of the runtime. (This is because we have to download several months worth of data to make corrections.)It would be great if we only have to download the data once for multiple forecasters. I think there are (at least) two ways to do this:
get_predictions()
so that it can take in a list of forecasters.get_predictions()
into two functions: one that gets the relevant data, and one that gets predictions with the data from the first function as input.The text was updated successfully, but these errors were encountered: