Home / Questions / 555395
Accepted
kev
Asked: 2022-09-06 06:54:26 +0800 CST 2022-09-06 06:54:26 +0800 CST

How to get looping regressors from regressions in R?

  • 772

I have the following DataFrame

enter image description here

I run the regressions using ir_total as the dependent variable and 3-way combinations of municipalities as regressor variables.

combinaciones <- apply(combn(independientes, m=3), 2, 
                      FUN=function(x){paste0("`", x, "`", collapse = ' + ')})
combinaciones

combinaciones <- sapply(paste("ir_total ~", combinaciones), as.formula)
modelos<-lapply(combinaciones, lm, df)

lapply(modelos, FUN=function(x) {cbind(r2=summary(x)$r.squared,
                                      coef1=coef(x)[2],
                                      coef2=coef(x)[3],
                                      coef3=coef(x)[4])})->datos


datos <- do.call(rbind, datos)
rownames(datos) <- NULL

datos

The result I get is:

enter image description here

But I hope to get a column with the combinations of each regression and be able to identify them since I need to manipulate them datosby applying the sort() function.

r

1 Answers

  1. Best Answer

    It is possible to extract the independent terms from a regression model obtained with lmthe following way:

    atributes(modelo$terms)$term.labels

    In this case, we can also obtain them using the function coef()as follows, omitting the first coefficient that corresponds to the intercept:

    names(coef(x))[-1]

    In this way, we can make the following modification to obtain the name of all the independent variables in the same column.

    lapply(modelos, FUN = function(x) {
  cbind(
    r2 = summary(x)$r.squared, 
    coef1 = coef(x)[2], 
    coef2 = coef(x)[3],
    coef3 = coef(x)[4], 
    terminos = paste(attributes(x$terms)$term.labels, collapse = " ")
  )
})

    Since in this case we know that we will always have three independent variables, we can get the name of each term in its own column.

    lapply(modelos, FUN = function(x) {
  cbind(
    r2 = summary(x)$r.squared, 
    coef1 = coef(x)[2], 
    coef2 = coef(x)[3],
    coef3 = coef(x)[4], 
    term1 = attributes(x$terms)$term.labels[1],
    term2 = attributes(x$terms)$term.labels[2],
    term3 = attributes(x$terms)$term.labels[3]
  )
})
    • 1