BiostatsUHNplus (version 1.0.2)

caterpillar_plot: Caterpillar plot. Useful for plotting random effects from hierarchical models, such as MCMCglmm::MCMCglmm() object, that have binary outcome.

Description

Caterpillar plot. Useful for plotting random effects from hierarchical models, such as MCMCglmm::MCMCglmm() object, that have binary outcome.

Usage

caterpillar_plot(
  subjID,
  subjLabel = NULL,
  remove.text.subjID = FALSE,
  mcmcglmm_object,
  orig_dataset,
  binaryOutcomeVar,
  prob = NULL,
  title = NULL,
  no.title = FALSE,
  subtitle = NULL,
  ncol = NULL,
  fonts = NULL,
  columnTextWidth = NULL,
  break.label.summary = FALSE
)

Value

ggplot object of caterpillar plot

Arguments

subjID

key identifier field for participant ID in data sets

subjLabel

text label field in dataset to replace key identifier field for participant ID with in plot (if provided)

remove.text.subjID

boolean indicating if non-numeric text should be removed from subjID in plot label. Note that this can only be used if there are non-duplicate participant IDs when non-numeric text is removed. Default is FALSE (if provided)

mcmcglmm_object

MCMCglmm model output

orig_dataset

data frame supplied to MCMCglmm function

binaryOutcomeVar

name of binary variable (0,1) that denotes outcome in MCMCglmm model

prob

probability for highest posterior density interval, similar to a confidence interval. Default is 0.95 (if provided)

title

title of the plot. Overrides default title (if provided)

no.title

boolean that denotes if title should be outputted in plot. Default is TRUE (if provided)

subtitle

subtitle of the plot. Overrides default subtitle (if provided)

ncol

number of columns in plot. Default is 2 (if provided)

fonts

character text that denotes font for title, subtitle, category labels, x-axis plot labels (if provided)

columnTextWidth

numeric that denotes character width for label text before breaking to start new line. Default is 20 characters (if provided)

break.label.summary

boolean to indicate if new line should start in label before (n, event) summary. Default is FALSE

Examples

data("ae");

ae$G3Plus <- 0;
ae$G3Plus[ae$AE_SEV_GD %in% c("3", "4", "5")] <- 1;
ae$Drug_1_Attribution <- 0;
ae$Drug_1_Attribution[ae$CTC_AE_ATTR_SCALE %in% c("Definite", "Probable", "Possible")] <- 1;
ae$Drug_2_Attribution <- 0;
ae$Drug_2_Attribution[ae$CTC_AE_ATTR_SCALE_1 %in% c("Definite", "Probable", "Possible")] <- 1;

prior2RE <- list(R = list(V = diag(1), fix = 1),
  G=list(G1=list(V=1, nu=0.02), G2=list(V=1, nu=0.02)));
  
model1 <- MCMCglmm::MCMCglmm(G3Plus ~ Drug_1_Attribution + Drug_2_Attribution, 
  random=~Subject + ae_category, family="categorical", data=ae, saveX=TRUE, 
  verbose=FALSE, burnin=2000, nitt=10000, thin=10, pr=TRUE, prior=prior2RE);
  
p <- caterpillar_plot(subjID = "Subject",
  mcmcglmm_object = model1,
  prob = 0.99,
  orig_dataset = ae,
  binaryOutcomeVar = "G3Plus")
  
p <- caterpillar_plot(subjID = "ae_category",
  mcmcglmm_object = model1,
  prob = 0.95,
  orig_dataset = ae,
  remove.text.subjID = FALSE,
  ncol = 4,
  binaryOutcomeVar = "G3Plus",
  subtitle = "System organ class (n, event)",
  title = "Odds Ratio for G3+ Severity with 95% Highest Posterior Density Interval",
  fonts = c("Arial", "Arial", "Arial", "Arial"),
  break.label.summary = TRUE)