ggplot绘图需求：按Pathway2分组连接Treatment组均值点

我拥有分组变量Pathway2和次级分组变量Treatment，希望在ggplot中按Pathway2分组，将Treatment两个水平对应的均值点用线连接。但当前代码生成的图形不符合预期：

• 当前效果：同一Treatment水平的点沿着Pathway2被线连接，而非每个Pathway2分组下的两个Treatment点之间连线。
• 目标效果：每个Pathway2分组内，两个Treatment水平的均值点用线连接，不同Pathway2间的线互不关联。

当前使用代码：

ggplot(df, aes(x = Pathway2, y = Detected.spots, color = Treatment)) +
    geom_point(size = 3, stat = 'summary', fun.y = 'mean', position = position_dodge(width = 0.5)) +
    geom_errorbar(stat = 'summary', fun.data = 'mean_se', width = 0, fun.args = list(mult = 1.96), 
                  position = position_dodge(width = 0.5)) +
    geom_line(stat = 'summary', aes(group = Treatment), position = position_dodge(width = 0.5)) +  # Összekapcsolt vonalak hozzáadása
    facet_wrap(~Cancer.type.in.co.culture + Time, ncol = 2) +
    labs(x = "Pathway", y = "Detected spot", color = "Treatment") +
    scale_color_manual(values = c("black", "red")) +
    theme_gray(base_size = 20)

方法1：直接调整geom_line的分组逻辑 #

问题核心是geom_line的group参数设置错误，将其改为Pathway2即可让ggplot在同一Pathway2分组内连接两个Treatment的均值点，同时保持所有几何对象的position_dodge参数一致：

ggplot(df, aes(x = Pathway2, y = Detected.spots, color = Treatment)) +
    geom_point(size = 3, stat = 'summary', fun = 'mean', position = position_dodge(width = 0.5)) +
    geom_errorbar(stat = 'summary', fun.data = 'mean_se', width = 0, fun.args = list(mult = 1.96), 
                  position = position_dodge(width = 0.5)) +
    # 修改group为Pathway2，实现同一Pathway下的Treatment点连线
    geom_line(stat = 'summary', fun = 'mean', aes(group = Pathway2), position = position_dodge(width = 0.5)) +
    facet_wrap(~Cancer.type.in.co.culture + Time, ncol = 2) +
    labs(x = "Pathway", y = "Detected spot", color = "Treatment") +
    scale_color_manual(values = c("black", "red")) +
    theme_gray(base_size = 20)

方法2：提前预处理数据（更推荐） #

先计算各分组的均值与95%置信区间，再用处理后的数据绘图，逻辑更清晰，避免stat参数带来的潜在问题：

library(dplyr)
library(ggplot2)

# 预处理：计算每个分组的均值、标准误和95%置信区间
summary_df <- df %>%
  group_by(Pathway2, Treatment, Cancer.type.in.co.culture, Time) %>%
  summarise(
    mean_spot = mean(Detected.spots, na.rm = TRUE),
    se_spot = sd(Detected.spots, na.rm = TRUE)/sqrt(n()),
    ci_low = mean_spot - 1.96 * se_spot,
    ci_high = mean_spot + 1.96 * se_spot
  ) %>%
  ungroup()

# 绘制目标图形
ggplot(summary_df, aes(x = Pathway2, y = mean_spot, color = Treatment)) +
    geom_point(size = 3, position = position_dodge(width = 0.5)) +
    geom_errorbar(aes(ymin = ci_low, ymax = ci_high), width = 0, position = position_dodge(width = 0.5)) +
    geom_line(aes(group = Pathway2), position = position_dodge(width = 0.5)) +
    facet_wrap(~Cancer.type.in.co.culture + Time, ncol = 2) +
    labs(x = "Pathway", y = "Detected spot", color = "Treatment") +
    scale_color_manual(values = c("black", "red")) +
    theme_gray(base_size = 20)

关键调整说明 #

• 原代码中group = Treatment会让ggplot沿x轴（Pathway2）连接同一Treatment的所有点，与需求相反；改为group = Pathway2后，将在每个Pathway分组内连接两个Treatment的均值点。
• 所有几何对象（点、误差棒、线）的position_dodge(width = 0.5)参数必须保持一致，确保线能精准对应到对应的均值点。

内容的提问来源于stack exchange，提问作者Mátyás Bukva