growfunctions (version 0.1)

fit_compare: Side-by-side plot panels that compare latent function values to data for different estimation models

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

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.

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(2009:2013))]

## 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 = 7,
                                n.burn = 3,
                                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 = 35,
                                     n.burn = 15,
                                     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 values",
                                units_name = "state",
                                units_label = cps$st)
}