R语言正弦曲线拟合异常修复及ggplot2实现24小时全周期拟合求助

• 拟合线出现跳转是因为原始数据Time列是无序状态（顺序为17、9、20、12），R基础绘图的lines()函数会严格按照输入点的顺序连线，不会自动按横坐标值排序
• 要覆盖完整24小时周期，需要生成连续的1-24小时时间序列做预测，不能仅使用原始4个观测点的时间计算预测值

Test_case <- structure(list(TEST_RESULT = c(1.31, 65.44, 8.59, 9.91), Time = c(17, 
9, 20, 12)), row.names = c(NA, -4L), class = c("tbl_df", "tbl", 
"data.frame"))

# 拟合正弦模型
xc <- cos(2*pi*Test_case$Time/24)
xs <- sin(2*pi*Test_case$Time/24)
fit.lm <- lm(Test_case$TEST_RESULT ~ xc + xs)

# 生成24小时连续序列做预测，步长0.1保证曲线平滑
full_time <- seq(1, 24, by = 0.1)
full_xc <- cos(2*pi*full_time/24)
full_xs <- sin(2*pi*full_time/24)
full_pred <- predict(fit.lm, newdata = data.frame(xc = full_xc, xs = full_xs))

# 绘图
plot(TEST_RESULT ~ Time, data = Test_case, xlim = c(1,24), ylim = c(0,70), pch = 16)
lines(full_time, full_pred, col = "blue", lwd = 2)

library(ggplot2)

# 构造预测结果数据框
pred_df <- data.frame(
  Time = full_time,
  TEST_RESULT = full_pred
)

ggplot() +
  # 绘制原始观测点
  geom_point(data = Test_case, aes(x = Time, y = TEST_RESULT), size = 3) +
  # 绘制平滑拟合正弦曲线
  geom_line(data = pred_df, aes(x = Time, y = TEST_RESULT), color = "blue", linewidth = 1) +
  # 坐标轴配置
  scale_x_continuous(breaks = seq(0, 24, 4), limits = c(1, 24), name = "时间（小时）") +
  scale_y_continuous(limits = c(0, 70), name = "检测结果") +
  theme_bw()

内容的提问来源于stack exchange，提问作者Kreitz Gigs