ggplot 按正确顺序分组的置信区间

Question

我有这个数据集，但我正在努力解决 2 个问题。

数据集的快照是这样的：

Score<-c(-2, 3, 4, -1, 3, 4, 5, -1, 3, 5, -3, 3, 5, 1, -4, 5, -2, 
         1, 3, 4, -4, 2, -1, 3, 4, -2, 3, 4, -1, 3, 4, 5, -1, 3, 5, -3, 3, 5, 1, -4, 5, -2, 
         1, 3, 4, -4, 2, -1, 3, 4)

Group<-c( "S", "S", "A", "S", "A", "S", "A", "S", "S", "A", "S", "A", "S", "A", 
          "S", "S", "A", "S", "A", "S", "A", "S", "S", "A", "S", "S", "S", "A", "S", "A", "S", "A", "S", "S", "A", "S", "A", "S", "A", 
          "S", "S", "A", "S", "A", "S", "A", "S", "S", "A", "S"
           )

Scenerio_ID <-c(1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 
                6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20,
1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 
                6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)


CombinedTable<-data.frame(Score,Group, Scenerio_ID) 

我有这段代码可以生成我想要的图形类型：

CombinedTable %>%
  mutate(Scenerio_ID = as.factor(Scenerio_ID)) %>%
  arrange(Scenerio_ID) %>%
  ggplot(aes(x = factor(Scenerio_ID), y = avg_Score, color = ifelse(avg_Score > 0, "U", "T"))) +
  geom_point(size = 2) +
  geom_errorbar(aes(ymin = lci_Score, ymax = uci_Score), position = "dodge") +
  labs(x = "A or S Per Scenerio", y = "Score", color = "") +
  facet_wrap(~ Scenerio_ID, nrow = 1L, strip.position = "bottom") +
  theme_classic() +
  scale_color_manual(values = c("U" = "blue", "T" = "red"))

我正在尝试做两件事：

1. 我希望能够将所有场景放在同一个图中，但为了让场景有序，由于某种原因，x 轴上场景的顺序似乎是随机排列的。

2. 我想创建一个只有特定场景的图表（例如，4、10、16、17、19）。

非常感谢以上两点的任何帮助。我的尝试如下：

selected_scenarios <- c("4", "10", "16", "17", "19")

MeansCombinedTable %>%
  filter(Scenerio_ID %in% selected_scenarios) %>%
  ggplot(aes(x = Group, y = avg_Score)) +
  geom_errorbar(aes(ymin = lci_Score, ymax = uci_Score), position = "dodge") +
  labs(x = "", y = "Score") +
  facet_wrap(~ Scenerio_ID, ncol = 5, scales = "free_x") +
  theme_classic()+ scale_color_manual(values = c("U" = "blue", "T" = "red"))

添加了 dplyr summarize 的这段代码仍然不起作用：

  mutate(Scenerio_ID = as.factor(Scenerio_ID)) %>%
  group_by(Group, Scenerio_ID) %>%
  dplyr::summarise(avg_Score = mean(Score),
            lci_Score = mean(Score) - 1.96 * sd(Score)/sqrt(n()),
            uci_Score = mean(Score) +  1.96 * sd(Score)/sqrt(n()),
            col = ifelse(avg_Score > 0, "U", "T")) %>%
  ggplot(aes(Scenerio_ID, y = avg_Score, color = col)) +
  geom_point(size = 2) +
  geom_errorbar(aes(ymin = lci_Score, ymax = uci_Score), position = "dodge") +
  labs(x = "A or S Per Scenerio", y = "Score", color = "") +
  theme_classic() +
  scale_color_manual(values = c("U" = "blue", "T" = "red"))```

Answer 1

给定的数据，只要正确分组归纳应该没问题。显然没有足够的测量来获得样本数据中所有点的置信区间：

CombinedTable %>%
  mutate(Scenerio_ID = as.factor(Scenerio_ID)) %>%
  group_by(Group, Scenerio_ID) %>%
  summarize(avg_Score = mean(Score),
            lci_Score = mean(Score) - 1.96 * sd(Score)/sqrt(n()),
            uci_Score = mean(Score) +  1.96 * sd(Score)/sqrt(n()),
            col = ifelse(avg_Score > 0, "U", "T")) %>%
  ggplot(aes(Scenerio_ID, y = avg_Score, color = col)) +
  geom_point(size = 2) +
  geom_errorbar(aes(ymin = lci_Score, ymax = uci_Score), position = "dodge") +
  labs(x = "A or S Per Scenerio", y = "Score", color = "") +
  theme_classic() +
  scale_color_manual(values = c("U" = "blue", "T" = "red"))

如果你只想选择几个进行绘图，请使用

summarize

和

%in%

：

CombinedTable %>%
  mutate(Scenerio_ID = as.factor(Scenerio_ID)) %>%
  group_by(Group, Scenerio_ID) %>%
  summarize(avg_Score = mean(Score),
            lci_Score = mean(Score) - 1.96 * sd(Score)/sqrt(n()),
            uci_Score = mean(Score) +  1.96 * sd(Score)/sqrt(n()),
            col = ifelse(avg_Score > 0, "U", "T")) %>%
  filter(Scenerio_ID %in% c(1, 4, 5)) %>%
  ggplot(aes(Scenerio_ID, y = avg_Score, color = col)) +
  geom_point(size = 2) +
  geom_errorbar(aes(ymin = lci_Score, ymax = uci_Score), position = "dodge") +
  labs(x = "A or S Per Scenerio", y = "Score", color = "") +
  theme_classic() +
  scale_color_manual(values = c("U" = "blue", "T" = "red"))