在ggplot中绘制来自相同数据框的多个变量

Question

我有一个可以定义的数据集：

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（在每个ggplot上都有手动颜色和线型），但我是ggplot的新手并且之前没有这样做过。试图在StackOverflow上模拟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()

...与this question中一样，错误为Error in eval(expr, envir, enclos) : object 'value' not found。

但尝试回答this question ......

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"))

...给出错误geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic?的三个实例。我可以得到一些帮助吗？

Answer 1

实际上这是你真正想要的：

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，将group=variable退出将导致以下可怕消息：

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

得到以下特性：