矩阵转为ggplot2画图的数据格式

作者：ahworld
链接：转换矩阵为ggplot2的输入格式
来源：微信公众号-seqyuan
著作权归作者所有，任何形式的转载都请联系作者。

在使用ggplot2画图时，如果拿到的数据是N x N的，那么画图很不方便，需要转换成行列名称为新的列的标识

下面的代码提供了转换的方法

M1 <- data.frame(matrix(1:25,nrow=5,dimnames=list(c("R1","R2","R3","R4","R5"),c("S1","S2","S3","S4","S5"))))
M1$S1 <- c("A1","A1","A2","A2","A2")
M1$S2 <- c("B1","B1","B1","B2","B2")
M1

	S1	S2	S3	S4	S5
R1	A1	B1	11	16	21
R2	A1	B1	12	17	22
R3	A2	B1	13	18	23
R4	A2	B2	14	19	24
R5	A2	B2	15	20	25

rownames(M1) <- paste(rownames(M1), M1$S1, M1$S2, sep="-")
M2 <- subset(M1, select=c("S3","S4","S5"))
M2 <- as.data.frame(as.table(as.matrix(M2)))
M2

	Var1	Var2	Freq
1	R1-A1-B1	S3	11
2	R2-A1-B1	S3	12
3	R3-A2-B1	S3	13
4	R4-A2-B2	S3	14
5	R5-A2-B2	S3	15
6	R1-A1-B1	S4	16
7	R2-A1-B1	S4	17
8	R3-A2-B1	S4	18
9	R4-A2-B2	S4	19
10	R5-A2-B2	S4	20
11	R1-A1-B1	S5	21
12	R2-A1-B1	S5	22
13	R3-A2-B1	S5	23
14	R4-A2-B2	S5	24
15	R5-A2-B2	S5	25

library(stringr)
Var1 <- data.frame(str_split_fixed(M2$Var1, "-", 3))
M2 <- cbind(Var1, M2)
M2 <- M2[,-4]
M2

	X1	X2	X3	Var2	Freq
1	R1	A1	B1	S3	11
2	R2	A1	B1	S3	12
3	R3	A2	B1	S3	13
4	R4	A2	B2	S3	14
5	R5	A2	B2	S3	15
6	R1	A1	B1	S4	16
7	R2	A1	B1	S4	17
8	R3	A2	B1	S4	18
9	R4	A2	B2	S4	19
10	R5	A2	B2	S4	20
11	R1	A1	B1	S5	21
12	R2	A1	B1	S5	22
13	R3	A2	B1	S5	23
14	R4	A2	B2	S5	24
15	R5	A2	B2	S5	25

参考：

https://www.tutorialspoint.com/how-to-convert-a-matrix-to-a-data-frame-with-column-names-and-row-names-as-new-columns-in-r