ggplot2：具有连续数据垂直列的条形图

Question

我正在尝试制作一个函数来生成大量（垂直）条形图，以按四分位数为各个组生成收入均值。我希望我的四分位列在数据框中使用它们的列名单独命名。

这使得我的测试数据：

quart_nms <- c("1st", "2nd", "3rd", "4th")
tile_nms <- function(vec, nms){
  if (length(vec) != length(nms)) stop(
    "length vec ", length(vec), " must equal length nms ", length(nms))
  out <- data.frame(t(unlist(vec)))
  colnames(out)<- nms
  out
}
t1 <- tile_nms(1:4, quart_nms)

这是我的功能。我认为，我理解 ggplot2 需要此类图表的宽格式数据，但我怀疑我没有正确传递变量列表。它们不应该是 y 值吗？

bar_plot <- function(.dt, .tit="U.S. Personal Income Distribution by Quartile",
                     .sub = NULL, .xl = "Income Quartiles", .yl = "Mean Income", 
                     ...){
  ggplot(data =.dt) + 
    geom_col(mapping = aes(y = list(1st, 2nd, 3rd, 4th)), stat = identity) +
    labs(title = .tit, subtitle = .sub)
}

gr1 <- bar_plot(.dt = t1, .sub = "pop subgroup name")

Answer 1

不知道为什么你认为 ggplot2 需要宽格式的数据。从来都不是这样。一旦您转换成长格式并对绘图功能进行一些更改，一切都会正常工作。

这是来自您的代码，未更改：

quart_nms <- c("1st", "2nd", "3rd", "4th")
tile_nms <- function(vec, nms){
  if (length(vec) != length(nms)) stop(
    "length vec ", length(vec), " must equal length nms ", length(nms))
  out <- data.frame(t(unlist(vec)))
  colnames(out)<- nms
  out
}
t1 <- tile_nms(1:4, quart_nms)

以下是我的代码，基于你的：

# needed libraries
library(tidyr)
library(ggplot2)

# convert to long format
tlong <- gather(t1, quartile, value)
tlong$quartile <- factor(tlong$quartile, levels = quart_nms) # make factor and put levels in the right order

# plotting function
bar_plot <- function(.dt, .tit="U.S. Personal Income Distribution by Quartile",
                     .sub = NULL, .xl = "Income Quartiles", .yl = "Mean Income", 
                     ...){
  ggplot(data =.dt) + 
    geom_col(mapping = aes(x = quartile, y = value)) +
    labs(title = .tit, subtitle = .sub, x = .xl, y = .yl)
}

# call plotting function with long-format data
bar_plot(.dt = tlong, .sub = "pop subgroup name")