如何使用分类变量来控制绘图中的pch值

Question

我发现我需要使用一个函数来控制散点图中使用的字符，而不是只传入一个分类变量。这是为什么？

这是我尝试过的。

这是输入文件的样子

~/data $ head avgpm25.csv
"pm25","fips","region","longitude","latitude"
9.77118522614686,"01003","east",-87.74826,30.592781
9.99381725284814,"01027","east",-85.842858,33.26581
10.6886181012839,"01033","east",-87.72596,34.73148
11.3374236874237,"01049","east",-85.798919,34.459133
12.1197644686119,"01055","east",-86.032125,34.018597
10.8278048723489,"01069","east",-85.350387,31.189731
11.5839280137523,"01073","east",-86.82805,33.527872
11.2619958748527,"01089","east",-86.588226,34.73079
9.41442269955754,"01097","east",-88.139667,30.722256

这是R会话的成绩单

> pollution <- fread("~/Downloads/data//avgpm25.csv")
> str(pollution)
Classes ‘data.table’ and 'data.frame':  576 obs. of  5 variables:
 $ pm25     : num  9.77 9.99 10.69 11.34 12.12 ...
 $ fips     : chr  "01003" "01027" "01033" "01049" ...
 $ region   : chr  "east" "east" "east" "east" ...
 $ longitude: num  -87.7 -85.8 -87.7 -85.8 -86 ...
 $ latitude : num  30.6 33.3 34.7 34.5 34 ...
 - attr(*, ".internal.selfref")=<externalptr> 

请注意pollution$region是chr

> with(pollution, plot(pm25))  # this works, but all points are black
> with(pollution, plot(pm25, col = region)) # this fails
Error in plot.xy(xy, type, ...) : invalid color name 'east'
> with(pollution, plot(pm25, col = as.factor(region))) # this works, points of two region get painted in two different colors

这就是我提问的原因。为什么这不起作用

> with(pollution, plot(pm25, col = as.factor(region), pch = as.factor(region)))
Error in plot.xy(xy, type, ...) : invalid plotting symbol

但定义一个函数，然后将其传递给它！

> pch_func <- function(x) { ifelse (x == "east", 1, 2)}
> with(pollution, plot(pm25, col = as.factor(region), pch = pch_func(region)))

这也不起作用

> pollution <- mutate(pollution, region = as.factor(region))
> with(pollution, plot(pm25, col = region, pch = region))
Error in plot.xy(xy, type, ...) : invalid plotting symbol

这是为什么？任何帮助将不胜感激。

Answer 1

来自?par

  

pch 指定符号或单个字符的整数   用作绘图点的默认值。查看可能值的点   和他们的解释

因此，您应该转换yourfactorto integer：

 with(pollution, plot(pm25, col = as.factor(region), 
            pch = as.integer(region)))

数据示例：

dx<- read.table(text='pm25,fips,region,longitude,latitude
                9.77118522614686,01003,east,-87.74826,30.592781
                9.99381725284814,01027,east,-85.842858,33.26581
                10.6886181012839,01033,east,-87.72596,34.73148
                11.3374236874237,01049,east,-85.798919,34.459133
                12.1197644686119,01055,aaa,-86.032125,34.018597
                10.8278048723489,01069,east,-85.350387,31.189731
                11.5839280137523,01073,aaaa,-86.82805,33.527872
                11.2619958748527,01089,east,-86.588226,34.73079
                9.41442269955754,01097,aaaa,-88.139667,30.722256',header=TRUE,sep=',')


with(dx, plot(pm25, col = as.factor(region), pch = as.integer(region)))