fit_compare: Side-by-side plot panels that compare latent function values...

In growfunctions: Bayesian Non-Parametric Dependent Models for Time-Indexed Functional Data

Description

Uses as input the output object from the gpdpgrow() and gmrfdpgrow() functions.

Usage

fit_compare(
  objects,
  H = NULL,
  label.object = c("gp_rq", "gmrf_rw2"),
  units_name = "Observation_Unit",
  units_label = NULL,
  date_field = NULL,
  x.axis.label = NULL,
  y.axis.label = NULL
)

Arguments

objects	A list input where each element is a returned object from estimation with either of gpdpgrow or gmrfdpgrow or an object that contains true N x T matrix of true latent function values, f. This latter input is only needed if want to compare estimated to true latent function values.
H	An N x 1 with entries in 1,...,M of cluster assignments for the N units of y under a known clustering.
label.object	A character vector of length equal to objects that contains labels for each element of objects to be used in rendering comparison plots. Defaults to label.object = c("gp_rq","gmrf_rw2").
units_name	A character input that provides a label for the set of N observation units. Defaults to units_name = "Observation_Unit".
units_label	A vector of labels to apply to the observation units with length equal to the number of unique units. Defaults to sequential numeric values as input with data, y.
date_field	A vector of Date values for labeling the x-axis tick marks. Defaults to 1:T .
x.axis.label	Text label for x-axis. Defaults to "time".
y.axis.label	Text label for y-axis. Defaults to "function values".

Value

A list object containing the plot of estimated functions, faceted by cluster, and the associated data.frame object.

p.t	A ggplot2 plot object
map	A data.frame object that contains clustering structure of observation units.

Author(s)

Terrance Savitsky [e-mail]

See Also

gpdpgrow, gmrfdpgrow

Examples

{
library(growfunctions)

## load the monthly employment count data 
## for a collection of 
## U.S. states from the Current 
## Population Survey (cps)
data(cps)
## subselect the columns of N x T, y, 
## associated with 
## the years 2009 - 2013
## to examine the state level 
## employment levels 
## during the "great recession"
y_short <- cps$y[,(cps$yr_label %in% 
                 c(2010:2013))]

## run DP mixture of GP's to 
## estimate posterior distributions 
## for model parameters
## uses default setting of a 
## single "rational quadratic" 
## covariance formula
res_gp         <- gpdpgrow(
                     y = y_short, 
                     n.iter = 3, 
                     n.burn = 1, 
                     n.thin = 1, 
                     n.tune = 0)  
## 2 plots of estimated functions: 
## 1. faceted by cluster and fit;
## 2.  data for experimental units.
## for a group of randomly-selected 
## functions
fit_plots_gp   <- cluster_plot( 
 object = res_gp,  units_name = "state", 
 units_label = cps$st, single_unit = FALSE, 
 credible = TRUE )
                                   
## Run the DP mixture of iGMRF's to 
## estimate posterior 
## distributions for model parameters
## Under default 
## RW2(kappa) = order 2 trend 
## precision term
res_gmrf     <- gmrfdpgrow(y = y_short, 
                       n.iter = 13, 
                       n.burn = 4, 
                       n.thin = 1) 
                                     
## 2 plots of estimated functions: 
## 1. faceted by cluster and fit;
## 2.  data for experimental units.
## for a group of randomly-selected functions
fit_plots_gmrf   <- cluster_plot( object = res_gmrf, 
  units_name = "state", units_label = cps$st, 
  single_unit = FALSE, 
  credible = TRUE )                                    
                                     
## visual comparison of fit performance 
## between gpdpgrow() and gmrfdpgrow()
## or any two objects returned from any
## combination of these estimation
## functions
objects        <- vector("list",2)
objects[[1]]   <- res_gmrf
objects[[2]]   <- res_gp
label.object   <- c("gmrf_tr2","gp_rq")
## the map data.frame object 
## from fit_plots gp 
## includes a field that 
## identifies cluster assignments
## for each unit (or domain)
H        <- fit_plots_gp$map$cluster
fit_plot_compare_facet <- 
fit_compare( objects = objects, 
 H = H, label.object = label.object,
 y.axis.label = "normalized y",
 units_name = "state", units_label = cps$st)                                  
}

growfunctions documentation built on July 17, 2021, 1:08 a.m.