We should have a plot like ppc_ribbon() but that plots the raw draws instead of computing interval summaries (like ppc_dens_overlay() but values not densities, and and x axis predictor or time variable).
So for example, ppc_curve_overlay(y, yrep, x) would plot nrow(yrep) overlaid regression lines (at least they are draws of the regression line if yrep is from posterior_linpred() instead of posterior_predict()).
We could call it ppc_line_overlay() but yrep doesn’t have to come from a linear model.
jgabry
commented
4 years ago
We should have a plot like
ppc_ribbon()but that plots the raw draws instead of computing interval summaries (likeppc_dens_overlay()but values not densities, and and x axis predictor or time variable).So for example,
ppc_curve_overlay(y, yrep, x)would plotnrow(yrep)overlaid regression lines (at least they are draws of the regression line ifyrepis fromposterior_linpred()instead ofposterior_predict()).We could call it
ppc_line_overlay()butyrepdoesn’t have to come from a linear model.