Different graph results when using ggplot vs when using base R functions?

I am creating some simple graphs for a project in R. I got unexpected results when using the ggplot2 package, and so I defaulted back to base R. When recreating the plot in base R, I got the correct result. Now I am just wondering why different results were produced?

Initially, this was the code and result using ggplot (which is incorrect):

ggplot(KBUF, aes(x = TEMP, y = HGHT)) +
geom_line(aes(color = "Temperature"), size = 1) +
geom_line(aes(x = DEWP, color = "Dew Point"), size = 1) +
scale_color_manual(values = c("blue", "red")) +
labs(x = "Temperature (°C)", y = "Height")

The correct plot I made in base R

has the following code:

plot(KNKX$TEMP, KNKX$HGHT, type = "l", col = "blue", xlab = "Temperature (°C)", ylab = "Height")

lines(KNKX$DEWP, KNKX$HGHT, col = "red")

legend("topright", legend = c("Temperature", "Dew Point"), col = c("blue", "red"), lty = 1)

Why are they different?

geom_line takes all the (x, y) co-ordinates and orders them from smallest x value to largest x value, then draws a line through these points. If you want to draw a line that goes through the points in the order they appear you need to use geom_path, not geom_line

For example, suppose I have a data frame that has the co-ordinates of a circle ordered counter-clockwise:

df <- data.frame(x = sin(pi * seq(-1, 1, 0.01)),
                 y = cos(pi * seq(-1, 1, 0.01)))

If I plot this with geom_line, I get a mess, because the line joins all the points according to their x value:

library(ggplot2)

ggplot(df, aes(x, y)) + geom_line()

If I want to preserve ordering, I simply use geom_path instead of geom_line:

ggplot(df, aes(x, y)) + geom_path()