首页 文章

R：如何根据特定规则从数据框中选择数据，并将数据作为新列添加到现有数据框中

提问于
浏览 次
1

我有两个数据帧df1和df2 .

df1 <- data.frame(x1=c("A35", "A41", "A49"),
                  x2=c(8, 24, 33),
                  x3=c(15, 63, 54))

df2 <- data.frame(y1=c("A35", "A38", "A41", "A41", "A49"),
                  y2 = c(9, 20, 24, 32, 84))

我想根据以下三个条件从df2中选择行

（1）df2的y1与df1的x1相匹配;

（2）df1的y2 = d2的x2

（3）df2的y2 = df1的<x3

符合条件的数据将作为新列添加到df1 . 如果df1的行具有多个匹配，那么这些附加匹配也将作为新列添加 .

预期的结果是

data.frame(x1=c("A35", "A41", "A49"),
           x2=c(8, 24, 33),
           x3=c(15, 63, 54),
           z1 = c("A35", "A41", ""),
           z2 = c(9, 24,""),
           z3 = c("", "A41", ""),
           z4 = c("", 32, ""))

x1 x2 x3 z1 z2 z3 z4
A35 8 15 A35 9  
A41 24 63 A41 24 A41 32
A49 33 54

提前致谢！

r dataframe pattern-matching

2 回答

  • 0

    如果我正确理解你的问题，这应该有效：

    ### we use the matches to pick our values from df1
### we use our conditions to pick our values from df2
matches <- match(df2$y1,df1$x1)
matches <- matches[!is.na(matches)]
condition1 <- df2$y1 %in% df1$x1
condition2 <- df2$y2[condition1] >= df1$x2[matches]
condition3 <- df2$y2[condition1] <= df1$x3[matches]

### i create these tmp variables so you can see step by step
### what each line of code is doing
### here i am finding the values that meet all the conditions
### then i am pulling the associated y2 values
tmp <- data.frame(x1=df1$x1[matches],y2=df2$y2[condition1])
tmp <- tmp[condition2&condition3,]
tmp <- droplevels(tmp)

### now that we have the values we want
### we are organizing the data in the desired output you 
### specified. 
x <- split(tmp[-1], tmp[[1]])
tmp2 <- data.frame()
for(i in 1:length(x)){

  df <- data.frame(t(unlist(x[[i]], use.names=FALSE)))
  colnames(df) <- seq(1,nrow(x[[i]]))
  tmp2 <- rbind.fill(tmp2,df)

}
colnames(tmp2) <- paste(rep("z",ncol(tmp2)),1:ncol(tmp2),sep="")
res <- data.frame(df1[df1$x1 %in% names(x),],tmp2)
res <- rbind.fill(res,df1[!df1$x1 %in% names(x),])

>res
   x1 x2 x3 z1 z2
1 A35  8 15  9 NA
2 A41 24 63 24 32
3 A49 33 54 NA NA
    回复于
  • 0

    建议不要使用长度不等的数据帧，使用列表应该更好地用于此目的 .

    我创建了一个完成工作的代码，尽管我不确定它是最有效的方法 .

    首先，您需要比较两个数据帧的每一行 . 这可以使用apply函数中的apply函数来完成（基本上：对于df1中的每一行，与df2中的每一行进行比较）并返回匹配值及其索引 .

    这将存储在一个杂乱的列表中，其中包含不匹配的空元素 . 因此，在清理列表后，可以使用sapply函数将生成的匹配添加到df1的每个单独行 .

    df1 <- data.frame(x1=c("A35", "A41", "A49"),
              x2=c(8, 24, 33),
              x3=c(15, 63, 54))

df2 <- data.frame(y1=c("A35", "A38", "A41", "A41", "A49"),
                  y2 = c(9, 20, 24, 32, 84))

matches <- apply(df2,1,function(x) apply(df1,1,function(y) 
  if(x[1]==y[1]&&x[2]>=y[2]&&x[2]<=y[3]){
    c(which(df1==x[1]),x[1:2])
  }))
addedelem <- t(array(unlist(matches), dim=c(3,length(unlist(matches))/3)))
result <- sapply(1:length(df1$x1), function(x) (c(as.matrix(df1[x,]),t(addedelem[which(addedelem[,1]==x),2:3]))))

    结果列表是您正在寻找的 . 如果有必要，您可以将其重新转换为数据帧 .

    > result
[[1]]
[1] "A35" "8"   "15"  "A35" " 9" 

[[2]]
[1] "A41" "24"  "63"  "A41" "24"  "A41" "32" 

[[3]]
[1] "A49" "33"  "54"
    回复于

相关问题