拿R来画画（八）：面积图与堆积图

面积图

面积图主要用来展示数量随时间变化的情况，单一变量的面积图在数字起伏较大时视觉效果较好。

平平无奇面积图

library(ggplot2)
library(gcookbook)
# 太阳黑子数据集
sunspots

A Time Series: 235 × 12

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
1749	58.0	62.6	70.0	55.7	85.0	83.5	94.8	66.3	75.9	75.5	158.6	85.2
1750	73.3	75.9	89.2	88.3	90.0	100.0	85.4	103.0	91.2	65.7	63.3	75.4
1751	70.0	43.5	45.3	56.4	60.7	50.7	66.3	59.8	23.5	23.2	28.5	44.0
1752	35.0	50.0	71.0	59.3	59.7	39.6	78.4	29.3	27.1	46.6	37.6	40.0

# 将timeseries数据转换成dataframe数据
dss = data.frame(year = as.numeric(time(sunspot.year)), ss = as.numeric(sunspot.year))
head(dss)

A data.frame: 6 × 2

year	ss
<dbl>	<dbl>
1700	5
1701	11
1702	16
1703	23
1704	36
1705	58

ggplot(dss, aes(x = year, y = ss)) + 
geom_area()

精致优雅面积图

# color指定边框颜色，fill指定填充颜色，alpha指定透明度
ggplot(dss, aes(x = year, y = ss)) + 
geom_area(color = "green", fill = "green", alpha = 0.3)

正负分开堆积图

dss[100:200,]$ss = -dss[100:200,]$ss
dss$tag <- dss$ss >= 0
# rle用来捕获连续运行长度，比如
# dat <- c(1, 2, 2, 2, 3, 1, 4, 4, 1, 1)
# r <- rle(dat)
# r
#   lengths: int [1:6] 1 3 1 1 2 2
#   values : num [1:6] 1 2 3 1 4 1
cat.rle = rle(dss$tag)
# rep.int用来生成重复序列
# rep.int(1:5,2)
# 1 2 3 4 5 1 2 3 4 5
dss$group = rep.int(1:length(cat.rle$lengths), times=cat.rle$lengths)
# group的作用是把每一正负分段分为不同的组
ggplot(dss, aes(x = year, y = ss, fill = tag, group = group)) + 
geom_area(alpha = 0.3) + 
scale_fill_manual(values = c('green','red'), guide = FALSE)

# 如果仅使用fill自动分组会出现奇怪的图形，中间部分上方的红色是由于ggplot将tag中被FALSE分隔的两个TRUE看做同一group连接起来导致的
ggplot(dss, aes(x = year, y = ss, fill = tag)) + 
geom_area(alpha = 0.3) + 
scale_fill_manual(values = c('green','red'), guide = FALSE)

# 可以看到每一段连续的tag共享同一个group
dss[107:115,]

A data.frame: 9 × 4

	year	ss	tag	group
	<dbl>	<dbl>	<lgl>	<int>
107	1806	-28.1	FALSE	2
108	1807	-10.1	FALSE	2
109	1808	-8.1	FALSE	2
110	1809	-2.5	FALSE	2
111	1810	0.0	TRUE	3
112	1811	-1.4	FALSE	4
113	1812	-5.0	FALSE	4
114	1813	-12.2	FALSE	4
115	1814	-13.9	FALSE	4

堆积图

堆积图可以理解为操作对象扩展至多个变量的面积图。

平平无奇堆积图

head(uspopage)

A data.frame: 6 × 3

Year	AgeGroup	Thousands
<int>	<fct>	<int>
1900	<5	9181
1900	5-14	16966
1900	15-24	14951
1900	25-34	12161
1900	35-44	9273
1900	45-54	6437

ggplot(uspopage, aes(x = Year, y = Thousands, fill = AgeGroup)) + 
geom_area(color = 'black', size = 0.2, alpha = 0.3) + 
scale_fill_brewer(palette = "YlOrRd")

顺序逆转堆积图

library(plyr)
ggplot(uspopage, aes(x = Year, y = Thousands, fill = factor(uspopage$AgeGroup, levels = rev(levels(uspopage$AgeGroup))))) + 
geom_area(color = 'black', size = 0.2, alpha = 0.3) + 
scale_fill_brewer(palette = "YlOrRd", breaks = rev(levels(uspopage$AgeGroup))) + 
labs(fill="AgeGroup")

去除竖线堆积图

ggplot(uspopage, aes(x = Year, y = Thousands, fill = AgeGroup)) + 
geom_area(alpha = 0.3) + 
scale_fill_brewer(palette = "YlOrRd") + 
geom_line(position = 'stack', color = 'black', size = 0.2)

# 折腾半天我感觉还是没线的最好看
ggplot(uspopage, aes(x = Year, y = Thousands, fill = AgeGroup)) + 
geom_area(alpha = 0.3) + 
scale_fill_brewer(palette = "YlOrRd")

百分比比堆积图

# 计算百分比，首先按年份分成多个独立的dataframe，在每个frame内部计算present
uspopage_prop <- ddply(uspopage, "Year", transform, Persent = Thousands/sum(Thousands)*100)

ggplot(uspopage_prop, aes(x = Year, y = Persent, fill = AgeGroup)) + 
geom_area(alpha = 0.3) + 
scale_fill_brewer(palette = "YlOrRd")