search

sdtaylor / phenology_estimators

3 stars 0 forks source link

peak estimate errors #4

Closed sdtaylor closed 5 years ago

sdtaylor commented 5 years ago

They look really weird due to errors only being whole numbers

fig_3_population_peak_errors_old

sdtaylor commented 5 years ago

fixed by setting bw=1 in the density. https://github.com/sdtaylor/phenology_estimators/commit/6ef6332fe789617f8162165644dda92ee2e6b6ef

fig_3_population_peak_errors

sdtaylor commented 5 years ago

Since the figure is done in a function, bw=1 is applied to all figures. This doesn't change the other figures. ie here's the before and after for onset errors. The only noticeable difference is the mean midway curve gets a bit wonky. nbd.

Before: fig_1_population_onset_errors_old

With bw=1: fig_1_population_onset_errors