本节根据2020-2021年美国疾病死亡率数据来绘制玫瑰图，后台回复关键词2021-4-5获取数据及代码，喜欢的小伙伴欢迎关注我的公众号R语言数据分析指南
加载R包
library(tidyverse)
加载数据
deaths <- read.delim("deaths.xls",header = T,sep="\t")
数据过滤
dat <- deaths %>%
filter(state == "United States" & year == "2020") %>%
select(state,year, month, all_cause,
heart_disease,starts_with("covid")) %>%
mutate(covid = covid_other + covid_only,
all_other = all_cause - covid - heart_disease) %>%
select(-c(all_cause, covid_other,covid_only)) %>%
pivot_longer(cols = c("heart_disease","covid", "all_other"),
names_to = "cause",
values_to = "number") %>%
group_by(month) %>%
mutate(label_y = sum(number) + 40000,
month = factor(month.abb[month], levels = month.abb))
- filter() 按行进行数据筛选 & 且
- select() 取需要的列
- starts_with() 取以covid开头的列
- mutate() 添加新列
- select(-c(**)) 删除列
- group_by(month) 以month对数据进行分组
- pivot_longer() 宽表转长表
- pivot_longer()函数有三个主要的参数：
- cols，表示哪些列需要转换
- names_to，表示cols选取的这些列的名字，构成了新的一列
- values_to，表示cols选取的这些列的值，构成了新的一列
- coord_polar() 转换为极坐标
- month.abb() 将月份转换为缩写
my_months <- sample(1:12)
month.abb[my_months]
> my_months
[1] 7 3 11 1 8 10 2 4 6 9 5 12
> month.abb[my_months]
[1] "Jul" "Mar" "Nov" "Jan" "Aug" "Oct" "Feb" "Apr" "Jun" "Sep" "May"
[12] "Dec"
数据可视化
ggplot(dat, aes(x = month, y = number, fill = cause)) +
geom_col(color = "#5B5A5A",
width = 1) +
geom_text(aes(label = month, y = label_y),
family = "Deckhouse Regular") +
scale_fill_manual(name = NULL,
values = c("#87c0e6","#ffa0aa","#808080"),
labels = c("All other deaths","COVID-19","Heart disease")) +
theme_void() +
theme(legend.position = c(.5,.1),
legend.direction = "horizontal") +
coord_polar()
