在同一图中绘制每日累积降水量和每日每小时降水量

Question

我有这样的数据：

datetime	aws.pr_sum	aws.ta_2m	timestamp
2020-06-14 12:00:00	0.00	19.20000	1635171717
2020-06-14 13:00:00	0.85	19.73833	1635171717
2020-06-14 14:00:00	0.00	21.95667	1635171717
2020-06-14 15:00:00	0.00	21.71333	1635171717
2020-06-14 16:00:00	0.00	21.04667	1635171717
2020-06-14 17:00:00	0.00	20.66500	1635171717

正如标题所说，我想在同一张图中绘制每日累积降水量和每日每小时降水量。我做什么：

df <- df  %>% group_by(timestamp) %>%
                    mutate(aws_pcp_cumulative = cumsum(aws.pr_sum)) 
c3 <- c("inca" = "#6aa6fa", "AWS" = "#2ECBE9")

p3 = ggplot(df, aes(x=datetime)) +
geom_bar(stat = "identity", aes(x = datetime + shift/2, y = aws.pr_sum, color="AWS"), size=0.7, fill=NA, width=wd ) +
geom_line(data = df, aes(x = datetime, y = aws_pcp_cumulative, color =  "AWS"), size=1.2,alpha=0.2) +
geom_bar(stat = "identity", aes(x = datetime - shift/2, y = inca.RR.max, color="inca"), size=0.7, fill=NA, width=wd) +
geom_line(data = df, aes(x = datetime, y = inca_pcp_cumulative, color = "inca"), size=1.2,alpha=0.2) +
geom_text(aes(label=laws, x = datetime, y = aws.pr_sum, color="AWS"), vjust = -0.5, size=3) +
geom_text(aes(label=linca, x = datetime, y = inca.RR.max, color="inca"), vjust = -0.5, size=3) +
scale_x_datetime(breaks=brks, labels = lbls, date_labels="%d.%m %H:%M") +
  labs(x="",
       y= "",
       color="zrážky [mm/h]") +
  scale_color_manual(values = c3) +
  theme(
    panel.grid.major.x = element_blank(),
    legend.position = c(0.01,0.95),
    legend.justification = c("left", "top"),
    legend.box.just = "left",
    legend.margin = margin(6, 30, 6,6),
    legend.title = element_text(size = 12))  
 

这是结果：https://imgur.com/a/MYmJhy7

正如我们所见，整个期间都会累积总和，但我希望每天计算一次。有什么想法吗？

Answer 1

很遗憾，您的所有时间点都来自同一日期。 使用dplyr::group_by和sum可以得到累计值：

library(tidyverse)

data <-
  tribble(
    ~datetime, ~aws.pr_sum, ~aws.ta_2m, ~timestamp,
    "2020-06-14 12:00:00", 0, 19.2, 1635171717L,
    "2020-06-14 13:00:00", 0.85, 19.73833, 1635171717L,
    "2020-06-14 14:00:00", 0, 21.95667, 1635171717L,
    "2020-06-14 15:00:00", 0, 21.71333, 1635171717L,
    "2020-06-14 16:00:00", 0, 21.04667, 1635171717L,
    "2020-06-14 17:00:00", 0, 20.665, 1635171717L
  )
data
#> # A tibble: 6 x 4
#>   datetime            aws.pr_sum aws.ta_2m  timestamp
#>   <chr>                    <dbl>     <dbl>      <int>
#> 1 2020-06-14 12:00:00       0         19.2 1635171717
#> 2 2020-06-14 13:00:00       0.85      19.7 1635171717
#> 3 2020-06-14 14:00:00       0         22.0 1635171717
#> 4 2020-06-14 15:00:00       0         21.7 1635171717
#> 5 2020-06-14 16:00:00       0         21.0 1635171717
#> 6 2020-06-14 17:00:00       0         20.7 1635171717

data %>%
  mutate(
    datetime = datetime %>% as.POSIXct(),
    day = datetime %>% as.Date()
  ) %>%
  group_by(day) %>%
  mutate(
    aws.pr_su.day = sum(aws.pr_sum)
  ) %>%
  pivot_longer(c(aws.pr_sum, aws.pr_su.day)) %>%
  qplot(datetime, value, color = name, data = .)

^{由reprex package (v2.0.1) 于 2021 年 10 月 27 日创建}