CODEWITHSUNDEEP
rlinear-regression
Kirsten
Kirsten

Reputation: 18198

How do I extract the coefficient names from the lm coefficients?

I have the following code which displays some coefficients from lm

fit <-lm(Petal.Width ~ Petal.Length, data=iris) 
cf <-coef(summary(fit,complete = TRUE)) 
colnames(cf)[4] <- "pval"
cf<- data.frame(cf)
cf <-cf[cf$pval < 0.05,]
cf <-cf[order(-cf$pval), ]
head(cf)
cf[1,1]

I want to extract the names in the left column ie (intercept) and petal length. I thought I could use cf[1,1] but it shows the estimate

Upvotes: 0

Views: 2357

Answers (2)

semaphorism
semaphorism

Reputation: 866

The tidyverse solution would be to use broom:

library(broom)
tidy_fit <- tidy(fit)

Results:

# A tibble: 2 x 5
  term         estimate std.error statistic  p.value
  <chr>           <dbl>     <dbl>     <dbl>    <dbl>
1 (Intercept)    -0.363   0.0398      -9.13 4.70e-16
2 Petal.Length    0.416   0.00958     43.4  4.68e-86

Then it's easy to extract the components that you want and the resulting code is more readable, e.g. tidy_fit$term to get the list of variables ((Intercept) and Petal.Length).

Upvotes: 1

Ronak Shah
Ronak Shah

Reputation: 389315

Those are extracted using rownames :

fit <-lm(Petal.Width ~ Petal.Length, data=iris) 
cf <-coef(summary(fit,complete = TRUE)) 
rownames(cf)
#[1] "(Intercept)"  "Petal.Length"

Upvotes: 1

Related Questions