如何用melt和ggplot绘制分区间堆叠百分比条形图（含分组）

Hey there! Let's fix up your stacked bar chart to meet all your requirements. Here's a complete, working code example first, followed by breakdowns of each key step:

Complete Working Code #

library(ggplot2)
library(reshape2)
library(scales)

# Original data with added group variable (8 individuals split into 2 groups)
Data <- data.frame(
  LMX = c(1.92, 2.33, 3.52, 5.34, 6.07, 4.23, 3.45, 5.64),
  Thriving = c(4.33, 6.54, 6.13, 4.85, 4.26, 6.32, 5.63, 4.55),
  Wellbeing = c(1.92, 2.33, 3.52, 2.34, 4.07, 3.23, 3.45, 4.64),
  Group = rep(c("Group A", "Group B"), each = 4) # Adjust group labels to match your actual data
)
rownames(Data) <- 1:8

# Melt data while preserving the Group variable
Data_A <- melt(Data, id.vars = c("Group"), measure.vars = c("LMX", "Thriving", "Wellbeing"))

# Create numerical interval groups with specified ranges
Data_A$value_range <- cut(
  Data_A$value,
  breaks = c(0, 1.99, 3.99, 5.99, 7),
  labels = c("0-1.99", "2-3.99", "4-5.99", "6-7"),
  include.lowest = TRUE,
  ordered_result = TRUE # Ensure intervals are sorted in low-to-high order
)

# Sort x-axis variables by total value (adjust if you want a different sort logic)
Data_A$variable <- factor(
  Data_A$variable,
  levels = names(sort(colSums(Data[, c("LMX", "Thriving", "Wellbeing")])))
)

# Plot the flipped stacked bar chart
ggplot(Data_A, aes(x = variable, y = value, fill = value_range)) +
  geom_bar(position = position_fill(reverse = FALSE), stat = "identity") +
  scale_y_continuous(labels = percent_format()) +
  scale_fill_manual(
    values = c("0-1.99" = "yellow", "2-3.99" = "orange", "4-5.99" = "red", "6-7" = "green"),
    name = "Value Range"
  ) +
  coord_flip() +
  facet_grid(. ~ Group) # Use `Group ~ .` for vertical facets instead

Key Fixes & Explanations #

1. Preserve Group Variable During Data Reshaping #

• We added a Group column directly to the original data frame (tweak the labels to match your actual grouping).
• When using melt(), we set id.vars = c("Group") to keep this variable intact, so we can use facet_grid() later without losing group information.

2. Bin Values Into Specified Intervals #

• The cut() function lets us split values into your exact ranges:
  • breaks defines the boundary points of each interval
  • ordered_result = TRUE ensures intervals are treated as an ordered factor, so they stack in low-to-high order
  • include.lowest = TRUE makes sure values exactly equal to the lower bound (like 1.92) are included in the correct interval

3. Sort Variables (As Requested) #

• We converted variable to a factor with levels sorted by the total sum of each variable. This makes the flipped bars appear in low-to-high order based on their total values. If you want a different sort order, just adjust the levels argument here.

4. Correct Stacking & Coloring #

• position_fill(reverse = FALSE) keeps the interval stacking order consistent with your range sequence (flip to TRUE if you want the opposite order)
• scale_fill_manual() maps each interval to your specified colors, with a clear legend title
• coord_flip() converts the chart to horizontal bars as you requested

5. Facet by Group #

• facet_grid(. ~ Group) splits the chart into side-by-side facets for each group. Swap to facet_grid(Group ~ .) if you prefer vertical facets instead.

内容的提问来源于stack exchange，提问作者Andreas G.