在ggplot中绘制同一数据框中的多个变量

Question

我有一个可以这样定义的数据集：

Week <- c("2015_52", "2016_01", "2016_02", "2016_03", "2016_04")
y1 <- runif(5, 0, 1)
y2 <- runif(5, 0, 1)
y3 <- runif(5, 0, 1)
df <- data.frame(Week, y1, y2, y3)

我想在同一个 ggplot 上绘制所有三个 y 随着时间的推移（每个都有手动颜色和线型），但我是 ggplot 新手，以前不需要这样做。尝试模拟 StackOverflow 上类似问题的答案会导致错误。

例如...

library(ggplot2)
ggplot(df, aes(x = Week, y = value, colour = variable)) + 
  geom_line()

...正如在这个问题中给出的错误Error in eval(expr, envir, enclos) : object 'value' not found。

但尝试回答这个问题......

ggplot() + 
  geom_line(data = df, 
            aes(x = Week, y = y1, 
                color = "black", linetype = "solid")) + 
  geom_line(data = df, 
            aes(x = Week, y = y2, 
                color = "red", linetype = "solid")) +
  geom_line(data = df,
            aes(x = Week, y = y3, 
                color = "orange", linetype = "dashed"))

...给出了三个错误实例geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic?。我可以得到一些帮助吗？

Answer 1

事实上，我认为这才是你真正想要的：

library(ggplot2)
library(reshape2)

set.seed(123)
Week <- c("2015_52", "2016_01", "2016_02", "2016_03", "2016_04")
y1 <- runif(5, 0, 1)
y2 <- runif(5, 0, 1)
y3 <- runif(5, 0, 1)
df <- data.frame(Week, y1, y2, y3)

mdf <- melt(df,id.vars="Week")

ggplot(mdf, aes( x=Week, y=value, colour=variable, group=variable )) + 
  geom_line() +
  scale_color_manual(values=c("y1"="black","y2"="red","y3"="orange")) +
  scale_linetype_manual(values=c("y1"="solid","y2"="solid","y3"="dashed"))

请注意，忽略该group=variable选项将导致出现以下可怕的消息：

geom_path: Each group consists of only one observation. Do you need to adjust the group
aesthetic?

产量：