var.names: Extract Variable Names from '' Objects

View source: R/utilities.R

var.namesR Documentation

Extract Variable Names from Objects


This function extracts variable names from a object for use in specifying alternate variable names in \funlove.plot. Optionally, a file can be written for easy editing of names.


          file = NULL, 
          minimal = FALSE)



a object; the output of a call to \


the type of output desired. Can either be "df" for a data.frame or "vec" for a named vector. See "Value". The default is "vec" unless file is not NULL.


optional; a file name to save the output if type = "df". See \funwrite.csv, which calls. Must end in .csv.


whether the output should contain all variable names (i.e., all rows that appear the output of or just the unique base variables. See "Details".


The goal of the function is to make supplying new variable names to the var.names argument in \funlove.plot easier. Rather than manually creating a vector or data.frame with all the variable names that one desires to change, one can use var.names() to extract variable names from a object and edit the output. Importantly, the output can be saved to a CSV file, which can be easily edited and read back into R for use in love.plot(), as demonstrated in the Example.

When minimal = TRUE, only a minimal set of variables will be output. For example, if the variables analyzed in are age, race, and married, and int = TRUE in, many variables will appear in the output, including expansions of the factor variables, the polynomial terms, and the interactions. Rather than renaming all of these variables individually, one can rename just the three base variables, and all variables that arise from them will be accordingly renamed. Setting minimal = TRUE requests only these base variables.


If type = "vec", a character vector the the variable names both as the names and the entries.

If type = "df", a data.frame with two columns called "old" and "new", each with the variables as the entries.

If file is not NULL, the output will be returned invisibly.


Not all programs can properly read the Unicode characters for the polynomial terms when requested. These may appear strange in, e.g., Excel, but R will process the characters correctly.


data(lalonde, package = "cobalt")

b1 <- ~ age + race + married, data = lalonde,
             int = TRUE)
v1 <- var.names(b1, type = "vec", minimal = TRUE)
v1["age"] <- "Age (Years)"
v1["race"] <- "Race/Eth"
v1["married"] <- "Married"
love.plot(b1, var.names = v1)

## Not run: 
b2 <- ~ age + race + married + educ + nodegree +
              re74 + re75 + I(re74==0) + I(re75==0), 
              data = lalonde)
var.names(b2, file = "varnames.csv")

##Manually edit the CSV (e.g., in Excel), then save it.
v2 <- read.csv("varnames.csv")
love.plot(b2, var.names = v2)

## End(Not run)

cobalt documentation built on Nov. 4, 2022, 1:05 a.m.