CODEWITHSUNDEEP
rlm
Vallo Varik
Vallo Varik

Reputation: 187

Retain factor labels in linear model results

How does one retain labels in linear models? How to have the slope labelled Bar or xBar in the coefficient table below?

dat <-  data.frame(
  y = c(1:3, 10:12),
  x = gl(2, 3, labels = c('Foo', 'Bar'))
)

coef(summary(mod <- lm(y ~ x, dat)))
#             Estimate Std. Error   t value     Pr(>|t|)
# (Intercept)      6.5  0.4082483  15.92168 0.0000909619
# x1              -4.5  0.4082483 -11.02270 0.0003850677

Actual use case is more complicated, tens of linear models with tens of variables. I do realize that I could write my own function that both retrieves the labels and fits the model, but I wonder if I am missing something simpler.

Upvotes: 0

Views: 254

Answers (1)

DaveArmstrong
DaveArmstrong

Reputation: 21992

It already works like that for me:

dat <-  data.frame(
  y = c(1:3, 10:12),
  x = gl(2, 3, labels = c('Foo', 'Bar'))
)

coef(summary(mod <- lm(y ~ x, dat)))
#             Estimate Std. Error   t value     Pr(>|t|)
# (Intercept)        2  0.5773503  3.464102 0.0257214207
# xBar               9  0.8164966 11.022704 0.0003850677

Upvotes: 1

Related Questions