如何在ggplot中使用cut函数生成的分箱因子变量绘制直方图？

Hey there! I get what you're trying to do—recreate that quick base R histogram-style bar chart from your cut factor variable using ggplot2. Let's break this down step by step, since ggplot relies on data frames (unlike base R which handles standalone vectors directly).

The Core Idea #

When you run plot(var) on a factor, base R automatically counts the frequency of each bin and draws a bar chart. To replicate this in ggplot2, we just need to give it the factor in a data frame, then let it do the frequency counting (or do it ourselves explicitly).

Method 1: Let ggplot Handle Frequency Counting (Simplest) #

You don't need to pre-calculate anything—ggplot can count bin frequencies automatically. Just wrap your factor variable in a data frame, then map it to the x-axis:

For your iris example:

# First run your original binning code
qv <- round(quantile(iris$Sepal.Length))
d <- cut(iris$Sepal.Length, qv[!duplicated(qv)], include.lowest=TRUE)

# Load ggplot2
library(ggplot2)

# Recreate the plot
ggplot(data.frame(Sepal_Length_Bins = d), aes(x = Sepal_Length_Bins)) +
  geom_bar(fill = "steelblue", alpha = 0.8) +
  labs(x = "Sepal Length Bins", y = "Frequency") +
  theme_minimal()

For your standalone var factor:

var <- structure(c(1L, 2L, 2L, 3L, 1L, 1L, 1L, 1L, 3L, 2L), .Label = c("[0,3]", "(3,40]", "(40,738]"), class = "factor")

ggplot(data.frame(Variable_Bins = var), aes(x = Variable_Bins)) +
  geom_bar(fill = "coral", alpha = 0.8) +
  labs(x = "Variable Bins", y = "Frequency") +
  theme_bw()

Method 2: Pre-Calculate Frequencies (Explicit Control) #

If you want to mirror exactly what plot() does under the hood, use table() to count bin occurrences, convert that to a data frame, then plot with geom_bar(stat = "identity"):

# For the iris example
freq_df <- as.data.frame(table(d))

ggplot(freq_df, aes(x = d, y = Freq)) +
  geom_bar(stat = "identity", fill = "steelblue") +
  labs(x = "Sepal Length Bins", y = "Frequency") +
  theme_minimal()

Why This Works #

ggplot2 is designed to work with data frames, so even a single vector needs to be wrapped in one to use its mapping syntax. Both methods will produce a plot identical to plot(var)—the only difference is whether you handle the frequency counting yourself or let ggplot take care of it.

内容的提问来源于stack exchange，提问作者John Thomas