local_predictive_ability: Plots local predictive ability

In ooelrich/oscbvar: Stuff for Bayesian vector autoregressions

Description

Plots local predictive ability over a set of smoothing variables for one or more models. Options for univariate distibutions (lpdens vs each of the pooling variables) or bivariate (size of dots corresponds to lpdens, x and y axes corresponds to two different pooling variables).

Usage

local_predictive_ability(
  df_preds,
  df_pooling,
  univariate = TRUE,
  bivariate = TRUE
)

Arguments

df_preds	Data frame with predictions. Needs to contain a column named "t" that contains time point as well as a column called "lpdens." It also needs to have a column called "method." If there is more than one unique value in the method column (which corresponds to a multiplicity of models), predictions from different models will be given different colors.
df_pooling	Data frame with pooling variables, as well as a column names "t" that contains time point and corresponds to the time points in df_preds.
univariate	Boolean. Should univariate plots be returned? Defaults to true.
bivariate	Boolean. Should bivariate plots be returned? Defaults to true.

Details

Details go here.

Value

A list of (gg)plots.

ooelrich/oscbvar documentation built on Sept. 8, 2021, 3:31 p.m.