R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Forest plot for multiple models

forestplotRegrObj

R Documentation

Forest plot for multiple models

Description

Plot different model fits with similar variables in order to compare the model's estimates and confidence intervals. Each model is represented by a separate line on top of eachother and are therefore ideal for comparing different models. This extra appealing when you have lots of variables included in the models.

Usage

forestplotRegrObj(regr.obj, skip.variables, add.empty_row, order.regexps,
  order.addrows, box.default.size, rowname.fn, xlab, xlog, exp,
  estimate.txt = xlab, zero, get_box_size = fpBoxSize, ...)

fpBoxSize(p_values, variable_count, box.default.size, significant = 0.05)

Arguments

`regr.obj`	A regression model object. It should be of coxph, crr or glm class. Warning: The glm is not fully tested.
`skip.variables`	Which variables to use. The variables should be the names of the fit output and not the true output names if you're using the rowname_translate_function.
`add.empty_row`	Add empty rows. This can either be a vector or a list. When you have a vector the number indicates the row number where the empty row should be added, the format is: c(3, 5). If you give a list you have the option of specifying the name of the row, the format is: `list(list(3, "my rowname"), list(5, "my other rowname"))`. The rows will be added at the 3rd row and 5th row from the original position. Ie you don't have take into account that the 5:th row will be at the 6:th position after adding the 3rd row.
`order.regexps`	A regexp vector that searches for matches along the original rownames and reorders according to those.
`order.addrows`	If there are ordered groups then often you want empty rows that separate the different groups. Set this to true if you want to add these empty rows between groups.
`box.default.size`	The default box size
`rowname.fn`	A function that takes a rowname and sees if it needs beautifying. The function has only one parameter the coefficients name and should return a string or expression.
`xlab`	x-axis label
`xlog`	If TRUE, x-axis tick marks are to follow a logarithmic scale, e.g. for logistic regressoin (OR), survival estimates (HR), poisson regression etc. Note: This is an intentional break with the original `forestplot` function as I've found that exponentiated ticks/clips/zero effect are more difficult to for non-statisticians and there are sometimes issues with rounding the tick marks properly.
`exp`	Report in exponential form. Default true since the function was built for use with survival models.
`estimate.txt`	The text of the estimate, usually HR for hazard ratio, OR for odds ratio
`zero`	Indicates what is zero effect. For survival/logistic fits the zero is 1 while in most other cases it's 0.
`get_box_size`	A function for extracting the box sizes
`...`	Passed to `forestplot`
`p_values`	The p-values that will work as the foundation for the box size
`variable_count`	The number of variables
`significant`	Level of significance .05

Examples

org.par <- par("ask" = TRUE)

# simulated data to test 
set.seed(10)
ftime <- rexp(200)
fstatus <- sample(0:2,200,replace=TRUE)
cov <- data.frame(
		x1 = runif(200),
		x2 = runif(200),
		x3 = runif(200))

library(rms)
dd <- datadist(cov)
options(datadist="dd")

fit1 <- cph(Surv(ftime, fstatus == 1) ~ x1 + x2 + x3, data=cov)
fit2 <- cph(Surv(ftime, fstatus == 2) ~ x1 + x2 + x3, data=cov)

forestplotRegrObj (regr.obj = fit1, new_page=TRUE)

library(forestplot)
forestplotRegrObj (regr.obj = list(fit1, fit2),
                   legend = c("Status = 1", "Status = 2"), 
                   legend_args = fpLegend(title="Type of regression"),
                   new_page=TRUE)
  
modifyNameFunction <- function(x){
  if (x == "x1")
    return ("Covariate A")
  
  if (x == "x2")
    return (expression(paste("My ", beta[2])))
  
  return (x)
}

forestplotRegrObj (regr.obj = list(fit1, fit2), 
                   col=fpColors(box=c("darkblue", "darkred")),
                   variablesOfInterest.regexp = "(x2|x3)",
                   legend = c("First model", "Second model"),
                   legend_args = fpLegend(title = "Models"),
                   rowname.fn = modifyNameFunction, new_page=TRUE)
  
par(org.par)

Results