导读
在一幅图里展示箱、散点和折线。
一、模拟输入数据
set.seed(1995)
# 随机种子
data=matrix(abs(round(rnorm(20, mean=20, sd=5))), 5, 4)
# 随机正整数，5行，4列
colnames(data)=paste("指标", 1:4, sep=".")
# 列名
rownames(data)=paste("样品", 1:5, sep=".")
# 行名
data # 查看数据，如下：
指标.1 指标.2 指标.3 指标.4
样品.1 25 18 20 20
样品.2 18 20 19 8
样品.3 21 20 21 13
样品.4 22 19 25 7
样品.5 28 28 13 20
二、调整格式
library(reshape)
input=melt(data)
input # 结果如下：
Var1 Var2 value
1 样品.1 指标.1 25
2 样品.2 指标.1 18
3 样品.3 指标.1 21
4 样品.4 指标.1 22
5 样品.5 指标.1 28
6 样品.1 指标.2 18
7 样品.2 指标.2 20
8 样品.3 指标.2 20
9 样品.4 指标.2 19
10 样品.5 指标.2 28
11 样品.1 指标.3 20
12 样品.2 指标.3 19
13 样品.3 指标.3 21
14 样品.4 指标.3 25
15 样品.5 指标.3 13
16 样品.1 指标.4 20
17 样品.2 指标.4 8
18 样品.3 指标.4 13
19 样品.4 指标.4 7
20 样品.5 指标.4 20
三、ggplot画箱图
geom_boxplot()：箱图
ggplot(input, aes(x=Var2, y=value, fill=Var2)) +
geom_boxplot() +
labs(x="指标(x轴)", y="值", fill="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black'))
ggplot(input, aes(x=Var2, y=value, color=Var2)) +
geom_boxplot() +
labs(x="指标(x轴)", y="值", color="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black'))
四、添加散点
geom_point()：散点图
ggplot(input, aes(x=Var2, y=value, fill=Var2)) +
geom_boxplot() +
labs(x="指标(x轴)", y="值", color="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black')) +
geom_point(pch=21, fill="black", color="deepskyblue")
ggplot(input, aes(x=Var2, y=value, color=Var2)) +
geom_boxplot() +
labs(x="指标(x轴)", y="值", color="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black')) +
geom_point()
ggplot(input, aes(x=Var2, y=value)) +
geom_boxplot(aes(color=Var2)) +
labs(x="指标(x轴)", y="值", color="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black')) +
geom_dotplot(binaxis='y', stackdir='center', dotsize=0.7, color="white", aes(fill=Var2)) +
theme(legend.title=element_blank())
五、添加连线
geom_line()：折线图
ggplot(input, aes(x=Var2, y=value)) +
geom_boxplot(aes(color=Var2)) +
labs(x="指标(x轴)", y="值", color="指标(图例)") +
theme(panel.grid=element_blank(), panel.background=element_rect(fill='transparent', color='black')) +
geom_dotplot(binaxis='y', stackdir='center', dotsize=0.7, color="white", aes(fill=Var2)) +
theme(legend.title=element_blank()) +
geom_line(aes(group=Var1), linetype="dashed", col="skyblue")
实战1：
## length
length = read.table("length.txt", header=T, sep="\t")
## wilcox
uhgg = length[length$group=="UHGG",]$length
cgr2 = length[length$group=="CGR2",]$length
mean(uhgg)
mean(cgr2)
wilcox.test(uhgg, cgr2, paired = TRUE)
wilcox.test(uhgg, cgr2, paired = TRUE)$p.value
##
Title = "Wilcoxon p value < 2e-16"
p = ggplot(length, aes(x = group, y = length, color=group)) +
labs(x="", y="Genome length", title=Title) +
theme_classic() +
geom_boxplot(width=0.5, outlier.colour = NA) +
# 不显示离群点
geom_jitter(width = 0.1) +
# 添加散点
geom_line(aes(group=pair), color="gray", alpha=0.4) +
scale_color_manual(values = c('orangered3','deepskyblue3')) +
theme(panel.grid = element_line(colour = 'white')) +
theme(legend.position = "none") +
theme(text = element_text(family="serif")) +
theme(axis.title = element_text(size = 15),
axis.text = element_text(size = 15),
axis.line = element_line(size = 1),
axis.ticks = element_line(size = 1),
title = element_text(size = 12))
ggsave(p, filename="length.pdf")
实战2：
ggplot(data, aes(x=Group, y=Ratio)) +
geom_boxplot(outlier.colour = NA, width=0.8) +
theme_classic() +
geom_jitter(width = 0.2, aes(color=Family)) +
scale_color_manual(values = colors) +
labs(x="", y="Relvative ratio") +
theme(axis.title = element_text(size = 20),
axis.text = element_text(size = 17),
axis.line = element_line(size = 1),
axis.ticks = element_line(size = 1),
title = element_text(size = 15)) +
theme(legend.title=element_text(face="bold"),
legend.text=element_text(size=rel(1.1))) +
scale_y_continuous(labels = scales :: percent)
一组不错的参数：
ggplot(alpha, aes(x=source, y=alpha, color=source)) +
geom_boxplot(width = 0.5,
size = 1,
outlier.colour = NA) +
# size粗细 width宽度 position_dodge间距
geom_jitter(width = 0.1, size = 3) +
# width点距 size点大小
theme_classic() +
scale_color_manual(values = c("Colon"="#00BA38",
"Stool"="#F8766D",
"Ileum"="#619CFF")) +
theme(legend.text=element_text(size=15),
legend.title=element_text(face='bold', size=20)) +
labs(x="",
y="Shannon index",
color="Source") +
theme(title = element_text(size = 15, face="bold"),
legend.position = "none") +
scale_y_continuous(expand = c(0, 0),
limits = c(0, 5)) +
theme(axis.title = element_text(size = 25),
axis.text.y = element_text(size = 18),
axis.text.x = element_text(size = 20),
axis.line = element_line(size = 1),
axis.ticks = element_line(size = 1))
参数汇总
geom_boxplot(width = 0.5,
size = 1,
outlier.colour = NA)
# 箱宽，线条粗细，不显示离群点
geom_jitter(width = 0.1,
size = 3)
# 抖动点，点距，点大小
参考：
https://www.cnblogs.com/ljhdo/p/4921588.html
添加两组比较显著性结果
R语言ggboxplot-一文掌握箱线图绘制所有细节
配对样本基因表达趋势：优化前后的散点连线图+拼图绘制
How to Connect Data Points on Boxplot with Lines?
