How to use _for_ in the dataframes in R programming

问题

I'm trying to fit several models in R using the for loop.

The data I want to fit is the classical Auto data in a R pre-build package. Let's discover the names of the columns of this dataframe:

names(Auto)

"mpg"          "cylinders"    "displacement" "horsepower"   "weight"       "acceleration" "year"        
"origin"       "name"

I want to fit the all these predictors one by one with the target 'mpg'.

Instead of doing:

autotest1 = lm(mpg~cylinders, data=Auto)

autotest2 = lm(mpg~displacement, data=Auto)

autotest3 = lm(mpg~horsepower, data=Auto)

autotest4 = lm(mpg~weight, data=Auto)

autotest5 = lm(mpg~acceleration, data=Auto)

autotest6 = lm(mpg~year, data=Auto)

autotest7 = lm(mpg~origin, data=Auto)

I'm trying to use the for loop:

for (var in names(Auto))
{

  cat(lm(mpg~var, data=Auto))

}

Error in model.frame.default(formula = mpg ~ var, data = Auto, drop.unused.levels = TRUE) : 
  variable lengths differ (found for 'var')

I'm also trying Auto[i] with the index i being the columns of Auto dataframe without any success. Anyone could help me?

回答1:

We can create a formula object with paste

for(var in names(Auto)) print(lm(paste('mpg ~', var), data = Auto))

Or with reformulate

for(var in names(Auto)) print(lm(reformulate(var, 'mpg'), data = Auto))

回答2:

Another option would be something like:

Auto <- mtcars
Auto$Var <- 1

Temp <- list()

for (i in 2:c(length(names(Auto))-1)) {

  print(names(Auto)[i])

  Auto$Var <- Auto[,i]

  ModelTemp <- lm(mpg~Var, data=Auto)
  Temp[[i-1]] <- ModelTemp
}

来源：https://stackoverflow.com/questions/62363528/how-to-use-for-in-the-dataframes-in-r-programming

标签

dataframe

for-loop