如何在R中使用字符串动态创建ggplot图表？

Ah, I've run into this exact issue before! The problem here is that ggplot2's aes() uses non-standard evaluation (NSE) — when you pass attr directly, it treats it as a literal column name instead of evaluating it to get the string stored in the variable. Here are a few clean ways to fix this:

1. Use the .data pronoun (tidyverse recommended) #

This is the most modern and readable approach. The .data pronoun lets you index columns using strings, just like you would with a regular data frame:

for (attr in names(dataset)) {
  # Create the plot
  plot <- ggplot(data = dataset, aes(x = .data[[attr]])) +
    geom_histogram()
  
  # Don't forget to print the plot! ggplot won't display it automatically in loops
  print(plot)
}

2. Use aes_string() (legacy method) #

Before the .data pronoun existed, aes_string() was the go-to for passing string column names. It still works, though it's less preferred now:

for (attr in names(dataset)) {
  plot <- ggplot(data = dataset, aes_string(x = attr)) +
    geom_histogram()
  print(plot)
}

3. Use quasiquotation with rlang #

If you're comfortable with tidy evaluation tools, you can convert the string to a symbol and inject it into aes() using !! (bang-bang):

library(rlang)

for (attr in names(dataset)) {
  plot <- ggplot(data = dataset, aes(x = !!sym(attr))) +
    geom_histogram()
  print(plot)
}

Quick note: Always print your plots in loops! #

One easy mistake here is forgetting to call print() on the plot object. When you run ggplot code interactively, it auto-prints, but in loops or functions, you have to explicitly print it to see the output.

内容的提问来源于stack exchange，提问作者Rahul