R中的隨機列循環

Question

我想構建一個數據框，其中有一系列包含另一列的隨機分配的列。數據有一些需要維護的結構。即我想隨機分配L許多時間，同時保持V的結構。我想要一個看起來像這樣的數據框;

L B V A 
1 1 1 2 10.9 
2 1 1 2 6.5 
3 1 1 2 8.6 
4 1 1 3 11.1 
5 1 1 4 13.1 
6 1 1 6 11.5 

並創建此;

 ID L B V A R1 R2 R3 R4 R5 
1 1_1_2 1 1 2 10.9 27 20 19 6 26 
2 1_1_2 1 1 2 6.5 27 20 19 6 26 
3 1_1_2 1 1 2 8.6 27 20 19 6 26 
4 1_1_3 1 1 3 11.1 6 28 4 26 26 
5 1_1_4 1 1 4 13.1 16 2 6 14 32 
6 1_1_6 1 1 6 11.5 17 21 3 11 25 

我可以做到這一點使用下面的腳本中手動，但我不知道是否有一個平穩的方式，使這個自動化的，因爲我想這樣做數百randomisations，使列R1，R2，R3。 。R n（所以這樣做的循環比手動重複代碼更可取）。

# Example Data Frame # 
df = data.frame(sample(1:33, 1000, replace = T), sample(1:3, 1000, replace = T), sample(1:9, 1000, replace = T), round(rnorm(1000, 10, 2),1)) 
colnames(df) = c("L", "B", "V", "A") 
df = transform(df,id=as.numeric(factor(df$V))) 
df = data.frame(as.factor(df[,1]),as.factor(as.numeric(df[,2])),as.factor(df[,5]),as.numeric(df[,4])) 
colnames(df) = c("L","B","V","A") 
df = df[order(df$L, df$B, df$V),] 
rownames(df) = NULL 
head(df) 

# ID # 
df$ID = paste(df[,1], df[,2], df[,3], sep = "_") 
ID = unique(as.vector(df$ID)) 

# R1 # 
ID2 = data.frame(ID, sample(ID)); colnames(ID2) = c("ID","R1") 
df = merge(df, ID2) 
df$R1 = as.factor(do.call(rbind, strsplit(as.vector(df$R1), split="_"))[,1]) 

# R2 # 
ID2 = data.frame(ID, sample(ID)); colnames(ID2) = c("ID","R2") 
df = merge(df, ID2) 
df$R2 = as.factor(do.call(rbind, strsplit(as.vector(df$R2), split="_"))[,1]) 

# R3 # 
ID2 = data.frame(ID, sample(ID)); colnames(ID2) = c("ID","R3") 
df = merge(df, ID2) 
df$R3 = as.factor(do.call(rbind, strsplit(as.vector(df$R3), split="_"))[,1]) 

# R4 # 
ID2 = data.frame(ID, sample(ID)); colnames(ID2) = c("ID","R4") 
df = merge(df, ID2) 
df$R4 = as.factor(do.call(rbind, strsplit(as.vector(df$R4), split="_"))[,1]) 

# R5 # 
ID2 = data.frame(ID, sample(ID)); colnames(ID2) = c("ID","R5") 
df = merge(df, ID2) 
df$R5 = as.factor(do.call(rbind, strsplit(as.vector(df$R5), split="_"))[,1]) 

我怎樣才能創建一個循環，這將做到這一點在ň列數？

Answer 1

從上面的代碼中，我終於得到了一個答案繼：

# ID # 
df$ID = paste(df[,1], df[,2], df[,3], sep = "_") 
ID = unique(as.vector(df$ID)) 

n = 5 
Rs  = as.vector(rep(NA,n)) 

for(i in 1:n){ 
Rs[i] =  paste("R",i, sep = "") 
} 
Rs 

for(i in 1:n){ 
df[,5+i] = NA 
colnames(df)[5+i] = paste(Rs[i]) 
ID = unique(as.vector(df$ID)) 
ID2 = data.frame(ID, sample(ID)) 
ID2 = merge(df, ID2) 
df[5+i] = as.factor(do.call(rbind, strsplit(as.vector(ID2[,6+i]), split="_"))[,1]) 
} 
head(df) 

給出結果：

L B V A ID R1 R2 R3 R4 R5 
1 1 1 2 10.1 1_1_2 21 12 27 4 26 
2 1 1 4 7.7 1_1_4 7 29 2 9 10 
3 1 1 5 9.7 1_1_5 27 27 3 1 22 
4 1 1 5 8.3 1_1_5 27 27 3 1 22 
5 1 1 7 9.5 1_1_7 13 15 32 19 11 
6 1 1 7 12.4 1_1_7 13 15 32 19 11