强制facet_grid以与数据集中显示的顺序相同的顺序绘制构面 [英] Force facet_grid to plot facets in the same order as they appear in the data set

问题描述

我正在研究单个案例研究的视觉表示。我需要对ggplot2中的图形进行一些更改，但是我发现这有些挑战。这是我用来制作可重现示例的玩具数据集中包含的变量的简要说明：

I am working on visual representations of single case studies. I need to make some changes to my graph in ggplot2, but I found this to be a bit challenging. Here is a brief description of the variables contained in the toy data set that I used to make a reproducible example:

• 场合：会话评估程序评估了行为（从1到n）；


• 时间：每种情况的数量（基线从1到n，干预从1到n）；


• 阶段：条件（A =基线或B =干预）；


• ID：研究中的学生代码


• 结果：行为清单上的总分。

• Occasion: Number of the session rater evaluated the behavior (from 1 to n);
• Time: Number of each condition (baseline from 1 to n and intervention from 1 to n);
• Phase: Condition (A = baseline or B = intervention);
• ID: student code in the study
• outcome: total score on a behavioral checklist.

根据数据集中的标准（即第一次干预会话）对病例进行排序。不幸的是，当我用 ggplot2 :: facet_grid 创建了不同的构面时，案例按其编号排序，我在下图中看到了。我试图更改变量类型（从整数到因数，从因数到字符，等等），但是似乎没有什么改变。最后，由于实际数据集还包含其他几种情况，因此我无法手动订购这些方面。

The cases are ordered based on a criterion (i.e., the first intervention session) in the data set. Unfortunately, when I created different facets with ggplot2::facet_grid, the cases are sorted by their number and I got what you can see in the image below. I tried to change the variable type (from integer to factor, from factor to character, etc.), but nothing seemed to change. Finally, I can't order the facets manually because the real data set consists of several more cases.

outcome <- c(4, 8, 10, NA, 15, 7, 7, 9, 14, NA, 16, 4, 3, 2, 2, 7, 7, 9, 14, NA, 3, 6, 6, NA, 5, 9, 11, NA, 6, 3, 4, 8, 7, NA, NA, 3)
Phase <- c("A", "A", "B", "B", "B", "B", "B", "B", "B", "B", "B", "A", "A", "A", "B", "B", "B", "B", "B", "A", "A", "A", "A", "B", "B", "B", "B", "A", "A", "A", "A", "B", "B", "B", "B", "B")
Time <- c(1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 5)
Occasion <- c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8, 9)
ID <- c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 7, 7, 7, 7, 7, 7, 7, 7, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2)


db <- data.frame(ID, Occasion, Time, Phase, outcome)

intervention_lines <- db %>% 
  filter(Phase == "A") %>% 
  group_by(ID, Phase) %>%
  summarise(y = max(Occasion)) 

db %>% na.omit(outcome) %>% 
  ggplot(aes(x = Occasion, y = outcome, group = Phase)) + 
  geom_point(size = 1.8) + 
  geom_line(size = 0.65) +
  facet_grid(ID ~ .) +
  scale_x_continuous(name = "y", breaks = seq(0, 11, 1)) +
  scale_y_continuous(name = "x", limits = c(0, 30)) +
  theme_classic() +
  theme(strip.background = element_blank()) +
  annotate("segment", x = -Inf, xend = Inf, y = -Inf, yend = -Inf) +
  geom_vline(data = intervention_lines, aes(xintercept = y + 0.5), colour = "black", linetype = "dashed")

推荐答案

我遇到了一些麻烦，因为您的绘图在不同图层上使用了两个数据框，因此它们需要

I ran into some trouble since your plot is using two data frames for different layers, and they need matching factors for the facet ordering to work.

我通过将ID转换为一个因子，然后通过 intervention_lines $ y

I did that by converting ID to a factor and then ordering it by intervention_lines$y in both places.

library(forcats)
intervention_lines <- db %>% 
  filter(Phase == "A") %>% 
  group_by(ID, Phase) %>%
  summarise(y = max(Occasion)) %>%
  ungroup() %>%
  mutate(ID = ID %>% as_factor() %>% fct_reorder(y))

db %>% na.omit(outcome) %>% 
  mutate(ID = as_factor(ID)) %>%
  left_join(intervention_lines %>% select(ID, y)) %>%
  mutate(ID = ID %>% fct_reorder(y)) %>% 
  ggplot(aes(x = Occasion, y = outcome, group = Phase)) + 
  geom_point(size = 1.8) + 
  geom_line(size = 0.65) +
  scale_x_continuous(name = "y", breaks = seq(0, 11, 1)) +
  scale_y_continuous(name = "x", limits = c(0, 30)) +
  theme_classic() +
  theme(strip.background = element_blank()) +
  annotate("segment", x = -Inf, xend = Inf, y = -Inf, yend = -Inf) +
  geom_vline(data = intervention_lines, aes(xintercept = y + 0.5), colour = "black", linetype = "dashed") +
  facet_grid(ID~.)

< a href = https://i.stack.imgur.com/vmGLE.png rel = nofollow noreferrer>

这篇关于强制facet_grid以与数据集中显示的顺序相同的顺序绘制构面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！