R语言x轴为因子时xlim设置异常及时间范围限制解决方案

问题核心 #

你遇到的问题本质是：barplot()的x轴坐标不是直接对应因子的水平索引，而是函数内部计算的每个条形的中心位置。直接用因子水平数（360、1380）设置xlim，会和barplot实际使用的坐标系统不匹配，导致图形显示异常（变窄或超出窗口）。

解决方案 #

方案1：提前筛选目标时间范围的数据（最简便） #

直接从数据中过滤出6:00到23:00的行，再绘制条形图，无需手动设置xlim：

# 读取数据（可直接导入你提供的数据集）
dat <- read.csv("你的数据文件路径")

# 生成6:00到23:00的目标时间序列
target_times <- format(seq(as.POSIXct("2020-02-20 06:00:00"), 
                           as.POSIXct("2020-02-20 23:00:00"), 
                           by = "1 min"), "%H:%M")

# 绘图布局设置
par(mfrow = c(4, 1), mar = c(5.1, 2.1, 4.1, 2.1))

# 绘制A组（周一至周四）筛选后的数据
dat_A_filtered <- dat[dat$group == "A" & dat$time %in% target_times, ]
barplot(count ~ time, dat_A_filtered, ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Montag bis Donnerstag")

# 绘制B组（周五）筛选后的数据
dat_B_filtered <- dat[dat$group == "B" & dat$time %in% target_times, ]
barplot(count ~ time, dat_B_filtered, ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Freitag")

# 绘制C组（周六）筛选后的数据
dat_C_filtered <- dat[dat$group == "C" & dat$time %in% target_times, ]
barplot(count ~ time, dat_C_filtered, ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Samstag")

# 绘制D组（周日）筛选后的数据
dat_D_filtered <- dat[dat$group == "D" & dat$time %in% target_times, ]
barplot(count ~ time, dat_D_filtered, ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Sonntag")

方案2：基于barplot的条形位置设置xlim #

如果需要保留完整因子水平但仅显示指定区间，可以先获取所有条形的位置，再计算正确的xlim范围：

# 读取数据
dat <- read.csv("你的数据文件路径")

# 绘图布局设置
par(mfrow = c(4, 1), mar = c(5.1, 2.1, 4.1, 2.1))

# 处理A组：先获取完整条形位置，再计算目标xlim
full_pos_A <- barplot(count ~ time, dat[dat$group == "A", ], plot = FALSE)
bar_width <- diff(full_pos_A)[1] # 获取单个条形宽度
xlim_A <- c(full_pos_A[360] - bar_width/2, full_pos_A[1380] + bar_width/2)
barplot(count ~ time, dat[dat$group == "A", ], xlim = xlim_A, 
        ylim = c(0, max(dat$count)), col = "violet", border = NA, 
        xlab = "", ylab = "", yaxt = "n", main = "Montag bis Donnerstag")

# 处理B组：重复上述逻辑
full_pos_B <- barplot(count ~ time, dat[dat$group == "B", ], plot = FALSE)
xlim_B <- c(full_pos_B[360] - bar_width/2, full_pos_B[1380] + bar_width/2)
barplot(count ~ time, dat[dat$group == "B", ], xlim = xlim_B, 
        ylim = c(0, max(dat$count)), col = "violet", border = NA, 
        xlab = "", ylab = "", yaxt = "n", main = "Freitag")

# 处理C组（使用默认范围）
barplot(count ~ time, dat[dat$group == "C", ], ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Samstag")

# 处理D组（使用默认范围）
barplot(count ~ time, dat[dat$group == "D", ], ylim = c(0, max(dat$count)), 
        col = "violet", border = NA, xlab = "", ylab = "", yaxt = "n", 
        main = "Sonntag")

说明 #

• 方案1更直观，通过缩小数据范围让barplot自动适配x轴，彻底避免坐标不匹配问题。
• 方案2适合需要保留完整因子水平但仅展示部分区间的场景，核心是利用barplot(plot=FALSE)返回的条形位置来计算精准的xlim范围。

内容的提问来源于stack exchange，提问作者Ben