search

USGS-R / protoloads

Prototyping and exploring options for broad-scale load forecasting
0 stars 4 forks source link

Scale factor hack #71

Closed aappling-usgs closed 6 years ago

aappling-usgs commented 6 years ago

I think this PR gets us final numbers that can be used to generate plots. I also rebuilt the old plots to confirm that this PR fixes https://github.com/USGS-R/protoloads/issues/68.

aappling-usgs commented 6 years ago

The rebuilt plots:

inputs: fig_input_data

preds - loadest: fig_preds_v_time_loadest

preds - wrtds: fig_preds_v_time_wrtds

errors v lead time

jzwart commented 6 years ago

looks good but there are conflicts with indicator files and builds. I think because I built these and pushed in last PR. Yours are more recent builds though

aappling-usgs commented 6 years ago

I think I have the merge conflicts straightened out now. remake says everything is up to date for me (except 3_forecast.yml, which is intentional for now as noted in https://github.com/USGS-R/protoloads/issues/70)

aappling-usgs commented 6 years ago

PS - @jzwart I like your color scheme. Not gonna worry about carrying it over to the older plots quite yet, but I'm on board and we can work on unifying the plot styles toward that scheme later :+1:

jzwart commented 6 years ago

sounds good! we should also think about colors used for flow vs. model range vs. lead time