I'm not sure what causes this to happen yet, but there are some rows that show -Inf for nse_top10 and nse_bot10 metrics while the kge_top10 and kgebot10 (and rmse\<>) are either np.nan or have a finite value. Seems like the -Inf only appear for small number of observations.
I'm not sure what causes this to happen yet, but there are some rows that show -Inf for nse_top10 and nse_bot10 metrics while the kge_top10 and kgebot10 (and rmse\<>) are either np.nan or have a finite value. Seems like the -Inf only appear for small number of observations.