使用R中ggplot的facet_wrap的流动持续时间曲线 [英] Flow duration curve using facet_wrap of ggplot in R?

问题描述

我正在使用 hydroTSM软件包的 fdc .我有三个 data.frame ，我想使用 facet_wrap 功能构建 data.frame 的流动持续时间曲线(FDC)> ggplot 将 plots 放在三行和一列中.以下将为 DF1 生成 FDC曲线.

I am using fdc of hydroTSM package. I have three data.frame and i would like to construct Flow duration curves (FDC) of the data.frame using facet_wrap functionality of ggplot to have the plots in three rows and one column. the following will produce FDC curves for DF1.

library(tidyverse)
library(hydroTSM)
library(gridExtra)

DF1 = data.frame(Ob = runif(1000,0,500), A = runif(1000,0,700), B = runif(1000,2,800))
DF2 = data.frame(Ob = runif(1000,0,500), A = runif(1000,0,700), B = runif(1000,2,800))
DF3 = data.frame(Ob = runif(1000,0,500), A = runif(1000,0,700), B = runif(1000,2,800))

fdc(DF1, plot = TRUE)

我尝试将 gridExtra包与 grid.arrange 一起使用，以将三个图强制放在一个图形上.我不仅没有做到这一点，而且不是首选方法.我想使用 ggplot 的 facet_wrap 选项.实际上，使用 DF1 数据错误地绘制了该图.我正在寻找类似以下的内容:

I tried to use gridExtra package with grid.arrange to force the three plots on a single figure. I not only failed to do it but it is not the preferred method. I would want to used the facet_wrap options of the ggplot. In fact the figure is wrongly drawn by using DF1 data. i am looking for something like below:

更新:这是基于@Jon Spring的建议.

Update: This is based on @Jon Spring suggestions.

graphics.off()
rm(list = ls())

library(tidyverse)
library(hydroTSM)
library(gridExtra)

DF1 = data.frame(Ob = runif(800,0,500), M1= runif(800,0,700), M2 = runif(800,2,800), df = rep("Upstream", 800))
DF2 = data.frame(Ob = runif(1000,0,500), M1 = runif(1000,0,700), M2 = runif(1000,2,800), df = rep("Midstream", 1000))
DF3 = data.frame(Ob = runif(1000,0,500), M1 = runif(1000,0,700), M2 = runif(1000,2,800), df = rep("Downstream", 1000))

# combine data into one table with id column for the source
 bind_rows(DF1, DF2, DF3) %>% 
   # reshape into longer format
  pivot_longer(-df, names_to = "src", values_to = "flow") %>%
  arrange(-flow) %>%
  group_by(df, src) %>%
  mutate(flow_pct = 1 - percent_rank(flow)) %>%
  ungroup() %>%

  ggplot(aes(flow_pct, flow, color = src)) +
  geom_line() +
  theme_light() +
  facet_wrap(~df, ncol = 1) +
  labs(x = "% Time flow equalled or exceeded",
       y = "Q, [m3/s]") +
  theme(strip.text = element_text(hjust = 0, color = "black"),
        strip.background = element_blank())

推荐答案

您可以使用ggplot中的构面执行类似的操作:

You could do something like this with facets in ggplot:

library(tidyverse)
# combine data into one table with id column for the source
bind_rows(DF1, DF2, DF3, .id = "df") %>% 
  mutate(df = LETTERS[as.numeric(df)]) %>%
  # reshape into longer format
  pivot_longer(-df, names_to = "src", values_to = "flow") %>%
  arrange(-flow) %>%
  group_by(df, src) %>%
  mutate(flow_pct = 1 - percent_rank(flow)) %>%
  ungroup() %>%

  ggplot(aes(flow_pct, flow, color = src)) +
  geom_line() +
  theme_light() +
  facet_wrap(~df, ncol = 1) +
  labs(x = "% Time flow equalled or exceeded",
       y = "Q, [m3/s]") +
  theme(strip.text = element_text(hjust = 0, color = "black"),
        strip.background = element_blank())

如果要将字母注释放在更左端，则可以选择使用 patchwork 包来堆叠和标记图:

If you want the letter annotations placed farther left, you could alternately use the patchwork package to stack and label the plots:

library(tidyverse)
library(patchwork)

flow_plot <- function(df) {
  df %>% 
  pivot_longer(everything(), names_to = "src", values_to = "flow") %>%
  arrange(-flow) %>%
  group_by(src) %>%
  mutate(flow_pct = 1 - percent_rank(flow)) %>%
  ungroup() %>%

  ggplot(aes(flow_pct, flow, color = src)) +
  geom_line() +
  theme_light() +
  guides(color = guide_legend()) +
  labs(x = "% Time flow equalled or exceeded",
       y = "Q, [m3/s]") +
    theme(legend.position = c(0.85,0.6))
}


flow_plot(DF1) /
  flow_plot(DF2) /
  flow_plot(DF3) +
  plot_annotation(tag_levels = "A")

这篇关于使用R中ggplot的facet_wrap的流动持续时间曲线的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！