earth2mip.s2s package#

Submodules#

earth2mip.s2s.score module#

earth2mip.s2s.score.plot(df, output, timeUnit='yearmonthdate')#
earth2mip.s2s.score.score(terciles, truth)#

area-weighted RPSS scores

Regions: global, northern hemisphere, and southern hemisphere

Parameters:

  • terciles (Dataset) – predicted terciles. must have longitude, latitude, forecast_time coordinates/dims.

  • truth (Dataset) – true terciles. same format as terciles

Returns:

dataframe with scores for different regions and climatology. Example:

    lead_time forecast_time       t2m valid_time  week source         region        tp
0        21.0    2018-01-02  1.455134 2018-01-23   1.0   sfno         Global       NaN
1        35.0    2018-01-02  1.357457 2018-02-06   1.0   sfno         Global       NaN
2        21.0    2018-01-02  1.308716 2018-01-23   NaN   clim         Global  1.310107
3        35.0    2018-01-02  1.306281 2018-02-06   NaN   clim         Global  1.312259
4        21.0    2018-01-02  1.331612 2018-01-23   1.0   sfno  Northern Hem.       NaN
5        35.0    2018-01-02  1.211101 2018-02-06   1.0   sfno  Northern Hem.       NaN
6        21.0    2018-01-02  1.184829 2018-01-23   NaN   clim  Northern Hem.  1.237482
7        35.0    2018-01-02  1.180459 2018-02-06   NaN   clim  Northern Hem.  1.241959
8        21.0    2018-01-02  1.575785 2018-01-23   1.0   sfno  Southern Hem.       NaN
9        35.0    2018-01-02  1.497714 2018-02-06   1.0   sfno  Southern Hem.       NaN
10       21.0    2018-01-02  1.431765 2018-01-23   NaN   clim  Southern Hem.  1.381933
11       35.0    2018-01-02  1.430871 2018-02-06   NaN   clim  Southern Hem.  1.381818

Return type:

DataFrame

earth2mip.s2s.terciles module#

earth2mip.s2s.terciles.apply(path, tercile_edges, output)#
earth2mip.s2s.terciles.compute_edges(path, output)#

Module contents#