R - Timeseries, autoplot(), scale_y_continuous(), and the disappearing geom_abline()

[EnglishmanBob]

I'm attempting to fit a regression line to the some data (mens400 from the fpp2 library).

library(fpp2)
library(forecast)

m400 <- tslm(mens400 ~ tm400)

I'd like to produce a plot of the data with the regression line included.

autoplot(mens400) +
  geom_abline(slope = m400$coefficients[2], intercept = m400$coefficients[1],
              colour = "red")

Unfortunately everything gets squished. So I adjust the scale using scale_y_continuous:

autoplot(mens400) +
  scale_y_continuous(limits = c(40, 60)) +
  geom_abline(slope = m400$coefficients[2], intercept = m400$coefficients[1])

But, my regression line disappears.

The order of the objects doesn't seem to effect the outcome...

How do I get my regression line back?

[robjhyndman]

You haven't explained where tm400 comes from, but if it is equal to time(mens400) (which gives the vector of years beginning with 1896) the code will work.

library(fpp2)

m400 <- tslm(mens400 ~ time(mens400))

autoplot(mens400) +  
  geom_abline(slope = m400$coefficients[2], 
              intercept = m400$coefficients[1],
              colour = "red")

An alternative is just to use geom_smooth() like this:

autoplot(mens400) +
  geom_smooth(method='lm')
#> `geom_smooth()` using formula 'y ~ x'
#> Warning: Removed 3 rows containing non-finite values (stat_smooth).

^{Created on 2021-08-28 by the reprex package (v2.0.1)}

[technocrat]

@robjhyndman

geom_smooth(method='lm')

Footnote: add

geom_smooth(method='lm', se = FALSE)

to eliminate error band, if desired.