ggplot2热图未将值从高到低排列

Question

我正在尝试排列热图，以使具有较高“总体”的值出现在顶部，而具有较低“总体”的值出现在底部。问题在于数据（from collections import defaultdict d = defaultdict(list) for row in df.to_dict('records'): d[row.get('1')].append({row.get('0'): row.get('2')}) defaultdict(list, {'BBB': [{'2019-01-21': 11101110}, {'2019-01-20': 11101110}], 'CCC': [{'2019-01-21': 11101010}, {'2019-01-20': 11101010}], 'DDD': [{'2019-01-21': 10101110}, {'2019-01-20': 10101110}]}) ）的排列已经像这样，但是这种排列并未应用到最后。

有什么主意我可以解决这个问题吗？

谢谢。

mydata

^{由reprex package（v0.2.1.9000）于2019-01-21创建}

Answer 1

您需要获取希望track_name进入的顺序，然后使用该顺序创建一个因子。

library(tidyverse)
library(RColorBrewer)
library(extrafont)

plot_data <- mydata %>%
  gather(key, value, -c(track_name, album_name)) %>%
  filter(album_name == "The Money Store")

order <- plot_data %>%
  filter(key == "Overall") %>% 
  arrange(value) %>% 
  pull(track_name)

plot_data %>% 
  mutate(track_name = factor(track_name, order)) %>% 
  ggplot(aes(x = ordered(key, levels = c("Lyrics", "Sound", "Overall")), y = track_name, fill = value)) +
  geom_tile() + 
  coord_equal(expand = TRUE) + 
  labs(title = "How Angry Are Songs from Each Death Grips Album?", x = "", y = "", subtitle = "The Money Store (2012)") + 
  scale_fill_gradientn(colors = mypalette(100), breaks = c(60, 20), labels = c("More Angry", "Less Angry")) +
  guides(show.legend = FALSE) +
  theme_void() +
  theme(plot.title = element_text(family = "Arial", hjust = 0.5, size = 40, face = "bold", margin = margin(1, 0, 20, 0)), 
        axis.text.x = element_text(family = "Arial", size = 18, angle = 90, hjust = 1, face = "bold", margin = margin(6,0,3,0)),
        axis.text.y = element_text(family = "Arial", size = 18, face = "bold", margin = margin(0, 10, 0, 0)), 
        axis.ticks.y = element_blank(),
        axis.title = element_blank(), 
        legend.title = element_blank(),
        legend.text = element_text(family = "Arial", size = 14, face = "bold", margin = margin(1, 1, 1, 10)),
        legend.margin = margin(0, 0, 0, 30),
        plot.margin = unit(c(5, 0, 8, 0), "mm"),
        plot.subtitle = element_text(family = "Arial", hjust = 0.5, size = 22, face = "bold", margin = margin(0, 0, 30, 0)))

^{由reprex package（v0.2.1）于2019-01-21创建}