error while plotting multiple lines using ggplot

Question

I would like to plot four lines and the data is randomly generated below:

trE <- runif(12, 0.5, 0.8)
teE <- runif(12, 0.8, 1)
trES <- runif(12, 0, 0.3)
teES <- runif(12, 0.3, 0.5)
plotData <- data.frame(k=1:12, trE=trE, teE=teE, trES=trES, teES=teES)

I have plotted it using the below code:

ggplot(plotData, aes(k)) + 
  geom_line(aes(y = trE, colour = "Tr E")) + 
  geom_line(aes(y = teE, colour = "Te E")) +
  geom_line(aes(y = trES, colour = "Tr ES"), linetype="dashed") + 
  geom_line(aes(y = teES, colour = "Te ES"), linetype="dashed") +
  geom_hline(aes(yintercept = 0.5), linetype="dotted",colour='red') +
  scale_colour_manual(values=c("black", "orange","black", "orange"), 
                      labels=c("Tr E", "Te E", "Tr ES", "Te ES")) +
  scale_x_discrete(limits=1:12) +
  theme_bw()

But the output is not as expected:

The line colors are messed up and they should be in the order: "orange","black", "orange", "black" in the plot.

ggplot(plotData, aes(k)) + 
  geom_line(aes(y = trE, colour = "black")) + 
  geom_line(aes(y = teE, colour = "orange")) +
  geom_line(aes(y = trES, colour = "black"), linetype="dashed") + 
  geom_line(aes(y = teES, colour = "orange"), linetype="dashed") +
  geom_hline(aes(yintercept = 0.5), linetype="dotted",colour='red') +
  scale_colour_manual(values=c("black", "orange","black", "orange"), 
                      labels=c("Tr E", "Te E", "Tr ES", "Te ES")) +
  scale_x_discrete(limits=1:12) +
  theme_bw()

However, in this plot, the line colors are as expected but the labels are not as expected.

Do you have any thoughts for this weird behavior? or point me to the missing details.

update:

plotDataLong <- plotData %>% tidyr::gather(Error, value, 2:5)

ggplot(plotDataLong, aes(k, value, col=Error)) + geom_line() + 
  geom_hline(aes(yintercept = 0.5), linetype="dotted",colour='red') +
  scale_colour_manual(values=c("black", "orange","black", "orange"), 
                      labels=c("Tr E", "Te E", "Tr ES", "Te ES")) +
  scale_linetype(labels=c("solid","solid","dashed","dashed"))

Though the code is greatly simplified, the linetypes are not as expected.

Answer 1

As per @Richard's suggestion, I have tried using tidyr::gather and the expected solutions below.

plotDataLong <- plotData %>% tidyr::gather(Error, value, 2:5)

ggplot(plotDataLong, aes(k, value, col=Error,linetype=Error)) + geom_line() +
  geom_hline(aes(yintercept = 0.5), linetype="dotted",colour='red') +
  scale_colour_manual(values=c("black", "black", "orange", "orange"),
                      name="Errors",
                      breaks=c('teE', 'teES',  'trE', 'trES'),
                      labels=c("Te E", "Te ES", "Tr E", "Tr ES")) +
  scale_linetype_manual(labels=c("Te E", "Te ES", "Tr E", "Tr ES"),
                 name="Errors",
                 breaks=c('teE', 'teES',  'trE', 'trES'),
                 values=c("solid","dashed","solid","dashed"))

Answer 2

set.seed(1)
trE <- runif(12, 0.5, 0.8)
teE <- runif(12, 0.8, 1)
trES <- runif(12, 0, 0.3)
teES <- runif(12, 0.3, 0.5)
plotData <- data.frame(k=1:12, trE=trE, teE=teE, trES=trES, teES=teES)

library(ggplot2)
ggplot(plotData, aes(k)) + 
  geom_line(aes(y = trE, colour = "Tr E"),lwd=1) + 
  geom_line(aes(y = teE, colour = "Te E"),lwd=1) +
  geom_line(aes(y = trES, colour = "Tr ES"), linetype="dashed",lwd=1) + 
  geom_line(aes(y = teES, colour = "Te ES"), linetype="dashed",lwd=1) +
  geom_hline(aes(yintercept = 0.5), linetype="dotted",colour='red') +
  scale_colour_manual(
     values=c("Te E"="orange","Tr E"="black", "Te ES"="orange", "Tr ES"="black"),
     breaks=c("Te E", "Tr E", "Te ES", "Tr ES")) +
  scale_x_discrete(limits=1:12) +
  theme_bw() +
  guides(colour = guide_legend(keywidth = 2, 
    override.aes = list(linetype = c("solid", "solid", "dashed", "dashed"))))