ggplot2 swimplot包负y值问题：轴偏移与数据倾斜的解决方法

问题现象 #

当使用swimplot包绘制包含负起始年份（start_yr为负值）的泳道图时，会出现坐标轴偏移、泳道条对齐错误且显示倾斜的问题，这是因为swimmer_plot函数默认的坐标处理逻辑未适配负值场景。

核心原因 #

swimmer_plot内部默认以0为基准计算泳道条的位置，当存在负起始值时，函数会错误地偏移矩形元素的坐标，配合coord_flip后进一步加剧了错位问题。

解决方案：使用原生ggplot构建泳道图 #

绕过swimplot的封装限制，直接用ggplot的geom_rect手动绘制，完全控制坐标映射：

library(ggplot2)

df <- data.frame(study_id = c("ID1", "ID1", "ID1","ID1"),
                 start_yr = c(-2.3,0,0.05,1.14), end_yr = c(0,0.05,1.14,1.24),
                 treatment = c("S", "S+S", "Z+S", "Z+S"))

# 为每个study_id分配固定的y轴位置（多ID时可自动生成）
df$y_pos <- as.integer(factor(df$study_id))

ggplot(df, aes(y = y_pos)) +
  # 手动控制矩形的上下边界，对应原swimplot的width参数（0.85/2=0.425）
  geom_rect(aes(xmin = start_yr, xmax = end_yr, 
                ymin = y_pos - 0.425, ymax = y_pos + 0.425, 
                fill = treatment), color = NA) +
  coord_flip(ylim = range(df$y_pos) + c(-0.5, 0.5)) +  # 适配ID的y轴范围
  scale_x_continuous(breaks = seq(-3, 5, by = 0.5), limits = c(-3, 5)) +
  scale_y_continuous(breaks = unique(df$y_pos), labels = unique(df$study_id)) +
  theme_bw() +
  labs(y = "研究ID", x = "年份")

替代方案：修改swimmer_plot的输出图层 #

如果坚持使用swimplot，可以提取其生成的图层，修正坐标映射：

library(swimplot)
library(ggplot2)

df <- data.frame(study_id = c("ID1", "ID1", "ID1","ID1"),
                 start_yr = c(-2.3,0,0.05,1.14), end_yr = c(0,0.05,1.14,1.24),
                 treatment = c("S", "S+S", "Z+S", "Z+S"))

# 生成swimmer_plot对象
p <- swimmer_plot(df=df,id='study_id',
                  start="start_yr",end="end_yr", name_fill="treatment", 
                  width=0.85, color=NA)

# 修正图层中的xmin/xmax映射（替换默认的偏移逻辑）
p$layers[[1]]$mapping$xmin <- aes(xmin = start_yr)$xmin
p$layers[[1]]$mapping$xmax <- aes(xmax = end_yr)$xmax

# 添加坐标和刻度
p + coord_flip(clip = 'off', ylim = c(0.5,1.5)) + 
  scale_x_continuous(breaks=seq(-3,5,by=0.5), limits = c(-3,5)) +
  scale_y_continuous(breaks=1, labels="ID1")

内容的提问来源于stack exchange，提问作者Alex