search

bhlmn / kmlb-ts

Using ML classification methods to predict summertime thunderstorm initiation in Melbourne, FL
https://bhlmn.github.io/kmlb-ts/
MIT License
4 stars 1 forks source link

Add MOS as a reference method #4

Open bhlmn opened 7 years ago

bhlmn commented 7 years ago

Seems simple enough ... first order approximation would be to say that if t06.20 > 50%, then that means that MOS is forecasting a thunderstorm event. But the probabilities from previous runs aren't mutually exclusive ... how to reconcile this?

bhlmn commented 7 years ago

Could easily access information using row id ... http://stackoverflow.com/questions/16144739/get-data-frame-row-by-id

bhlmn commented 7 years ago

After thinking about it a little bit more, I think what makes the most sense is to: 

df.scaled$mos.prob <- max(t06.08, t06.14, t06.20)

If it is over 50%, count it as a hit.