将数据帧列值绘制为facet列

Question

我有以下形式的调查数据框架：

    category    shift   difficulty  importance  frequency   dsmImportance   supervisor
1   Monitoring  Day     3           1           1           3               Debra Smith
2   Monitoring  Day     2           1           1           3               Debra Smith
3   Paperwork   Night   3           1           1           3               Mark Hobbs
4   Operations  Day     1           1           1           2               Ryan Jones
5   Rostering   Night   1           1           1           1               Mark Hobbs

数据是对工作轮班期间执行的任务的调查，根据每项任务的难度、重要性等，对每项任务的评分为1-3。

我想要做的是绘制一个任务分级直方图数组，其中数组列为difficulty、importance、frequency和dsmImportance，行为category。到目前为止，我的方法是为每种评级类型(difficulty、importance等)创建单个列。使用category分面，然后使用grid_layout()将列组合在一起。您可以看到结果这里。(不幸的是，在我成为会员之前，我无法直接链接到一个图像。)它很管用，但它并不是很漂亮。

如何完全使用ggplot2**'s faceting函数创建数组？**我对R(和堆栈溢出)还不熟悉，但我很确定不能用它当前形式的数据来完成这个任务。我想我必须将数据熔化并转换成另一种形式，但我不知道应该是什么形式。

代码

library(ggplot2)
library(gridExtra)

walkaday.dirty = read.csv("~/Documents/walkaday.csv", header = TRUE, sep = ",", fill = TRUE, blank.lines.skip = TRUE)
walkaday = na.omit(walkaday.dirty)

// Order category levels by task frequency
category.levels = names(sort(table(walkaday$category), decreasing = TRUE))
walkaday$category = factor(walkaday$category, levels = category.levels)

难度图

difficulty = ggplot(walkaday, aes(factor(difficulty, c("3", "2", "1")), fill = difficulty)) + geom_bar() + coord_flip() + xlab("") + ylab("") + opts(legend.position = "none")
difficulty = difficulty + facet_grid(category ~ .) + opts(strip.text.y = theme_blank())

重要性图

importance = ggplot(walkaday, aes(factor(importance, c("3", "2", "1")), fill = importance)) + geom_bar() + coord_flip() + xlab("") + ylab("") + opts(legend.position = "none", axis.text.y = theme_blank(), axis.ticks = theme_blank())
importance = importance + facet_grid(category ~ .) + opts(strip.text.y = theme_blank())

频率图

frequency = ggplot(walkaday, aes(factor(frequency, c("3", "2", "1")), fill = frequency)) + geom_bar() + coord_flip() + xlab("") + ylab("") + opts(legend.position = "none", axis.text.y = theme_blank(), axis.ticks = theme_blank())
frequency = frequency + facet_grid(category ~ .) + opts(strip.text.y = theme_blank())

DSM重要性图

dsmImportance = ggplot(walkaday, aes(factor(dsmImportance, c("3", "2", "1")), fill = dsmImportance)) + geom_bar() + coord_flip() + xlab("") + ylab("") + opts(legend.position = "none", axis.text.y = theme_blank(), axis.ticks = theme_blank())
dsmImportance = dsmImportance + facet_grid(category ~ .) + opts(strip.text.y = theme_text(angle = 0))

组合图表

pushViewport(viewport(layout = grid.layout(1, 4, widths = c(1,1,1,1.7))))
print(difficulty + opts(title = "Task difficulty"), vp = viewport(layout.pos.row = 1, layout.pos.col = 1)) 
print(importance + opts(title = "Task importance"), vp = viewport(layout.pos.row = 1, layout.pos.col = 2)) 
print(frequency + opts(title = "Task frequency"), vp = viewport(layout.pos.row = 1, layout.pos.col = 3)) 
print(dsmImportance + opts(title = "DSM importance"), vp = viewport(layout.pos.row = 1, layout.pos.col = 4))

数据

可以找到数据集这里。

Answer 1

将数据转换为melt格式，以便评级类型显示为一个单独的变量：

walkaday <- read.csv("http://dl.dropbox.com/u/7046039/walkaday.csv")
walkaday.long <- melt(walkaday,id.vars=c(1,2,7))
qplot(factor(value,c("3","2","1")),data=walkaday.long,geom="bar")+facet_grid(.~variable)

注意，新变量的名称是variable，值是value。