Confidence Interval from LOESS dissapears when using weights with geom_smooth command

Question

In R, using the ggplot2 package, I have these two lines:

ggplot(data, aes(x = x, y = Pinera)) + 
    geom_point(shape = 1) + 
    geom_smooth(method = loess, se = TRUE)

ggplot(data, aes(x = x, y = Pinera, weight = pesos)) + 
    geom_point(shape = 1) + 
    geom_smooth(method = loess, se = TRUE)

The first one gives me an unweighted smooth curve from my data with the shaded confidence interval, however when I add the weights, the confidence interval region disappears despite I stated se = TRUE.

This is my data


Show source
| R   | ggplot2   | loess   2017-04-16 23:04 0 Answers

Answers ( 0 )

◀ Go back