如何从单个数据表/数据框在同一图上绘制两条线

Question

I have such data table structure, and I am so struggle how could I plot two line to clearly indicate two condition:

> Dependent   Counts     Indication
> 0           2          Stay   
> 1           4          Stay 
> 2           1          Stay
> 0           11         Left 
> 1           5          Left 
> 2           3          Left  
> 3           2          Left   
> 4           1          Left

[Btw] I saw some reference/ question asked and solution provided Dual line on single plot could definitely achievable by:

x <- c(...)
y1 <- c(...)
y2 <- c(...)

But this kind of solution did not hit my needs. I am seeking another way to do this as my table is prepared in the above-mentioned form.

Answer 1

if df is your dataframe:

df <- data.frame("Dependent"=c(0,1,2,0,1,2,3,4),
                 "Counts"=c(2,4,1,11,5,3,2,1),
                 "Indication"=c("Stay","Stay","Stay","Left","Left","Left","Left","Left"))

library(tidyr)
df_tidy <- spread(df, Indication, Counts)

plot(df_tidy$Dependent, df_tidy$Left, type="l")
lines(df_tidy$Dependent, df_tidy$Stay, col="red")

I am first using tidyr to "spread" your dataframe and from there you can use the standard plot function.

Spread:

This function takes data that is in a key-value format and returns a rectangular tidy cell format. As you can see spread() restructures the dataframe by removing redundant rows without losing any information.

For more information on spread and tidyr , please look here .

Answer 2

with ggplot

library(ggplot2)
ggplot(df, aes(x=Dependent, y=Counts, group=Indication, color=Indication)) + geom_line()

Answer 3

Finally I worked out the exact answer I am looking for.

ggplot() + geom_line(aes(subset(dataStay_Left$Dependents,
dataStay_Left$ind=="Stay"), subset(dataStay_Left$Counts, 
dataStay_Left$ind=="Stay"), group=1)) + 
geom_line(aes(subset(dataStay_Left$Dependents, dataStay_Left$ind=="Left"), 
subset(dataStay_Left$Counts, dataStay_Left$ind=="Left"), group=2))

Thanks for the help provided and the discussion.