如何让cpt.mean绘制的图像x轴显示Night值而非默认行号

cpt.mean()默认仅接收数值向量作为输入，自带的plot()方法默认使用向量的索引作为X轴刻度，不会关联你数据框中的Night字段。加上你的Night采样间隔不统一，不符合标准时间序列等间隔的要求，所以直接转时间序列的方法不可用，可选用以下方案解决：

快速修改方案 #

如果你可以接受X轴同一Night对应多个重复标签，只需要修改两行代码即可：

library(changepoint)
x <- cpt.mean(df$vocalisations, method="AMOC")
# 绘图时隐藏默认X轴
plot(x, xaxt = "n", xlab = "Night")
# 手动替换X轴刻度为Night值
axis(side = 1, at = seq_along(df$vocalisations), labels = df$Night)

无重复标签的自定义绘图方案 #

如果希望X轴仅展示一次每个Night的标签，手动重构绘图逻辑即可，完全匹配自带绘图的分段均值展示效果：

library(changepoint)
x <- cpt.mean(df$vocalisations, method="AMOC")
# 提取变点、分段信息
seg_ends <- c(cpts(x), nrow(df))
seg_starts <- c(1, cpts(x) + 1)
seg_means <- param.est(x)$mean

# 手动绘图
# 先绘制原始数据，X轴用Night值
plot(df$Night, df$vocalisations, type = "l", xlab = "Night", ylab = "vocalisations")
# 叠加分段均值水平线
for(i in seq_along(seg_means)){
  segments(
    x0 = df$Night[seg_starts[i]], y0 = seg_means[i],
    x1 = df$Night[seg_ends[i]], y1 = seg_means[i],
    col = "red", lwd = 2
  )
}
# 可选：添加变点位置垂直线标注
abline(v = df$Night[cpts(x)], col = "blue", lty = 2)

ggplot2 版本实现 #

需要更高自定义样式的可以用ggplot2绘制：

library(changepoint)
library(ggplot2)

x <- cpt.mean(df$vocalisations, method="AMOC")
# 生成分段数据框
seg_df <- data.frame(
  start_night = df$Night[c(1, cpts(x)+1)],
  end_night = df$Night[c(cpts(x), nrow(df))],
  mean_val = param.est(x)$mean
)

ggplot(df, aes(x = Night, y = vocalisations)) +
  geom_line() +
  geom_segment(
    data = seg_df, 
    aes(x = start_night, xend = end_night, y = mean_val, yend = mean_val),
    color = "red", linewidth = 1
  ) +
  geom_vline(xintercept = df$Night[cpts(x)], color = "blue", linetype = "dashed") +
  labs(x = "Night", y = "vocalisations") +
  theme_bw()

内容的提问来源于stack exchange，提问作者93harrison