Question

我有一组列名，如下所示

  "fb_metrics.3.name"                                     
  "fb_metrics.3.period"                                   
  "fb_metrics.3.values.0.value.share"                     
  "fb_metrics.3.values.0.value.like"                      
  "fb_metrics.3.values.0.value.comment"                   
  "fb_metrics.3.title"                                    
  "fb_metrics.3.description"

名称和期间每个只有一个唯一值。例如

> df[,"fb_metrics.3.name"]
[1] post_storytellers_by_action_type
Levels: post_storytellers_by_action_type

> df[, "fb_metrics.3.period"]
[1] lifetime
Levels: lifetime

我想像这样重命名列3,4,5

[1] "post_storytellers_by_action_type.lifetime.share"  
[2] "post_storytellers_by_action_type.lifetime.like"   
[3] "post_storytellers_by_action_type.lifetime.comment"

我已经像这样管理了替换位

i=3
new_column_name <- paste(as.character(df[1,paste("fb_metrics.",
                    i,".name", sep = "")]),as.character(df[1,paste("fb_metrics.",
                    i,".period", sep = "")]), sep = "." )

sub(pattern = ".*value",replacement = new_column_name,x = colnames(df[,
        grep(paste("fb_metrics.",i,".values.*",sep = ""), column_names)]))

我已经像这样提取了要替换的列

column_names <- colnames(df)

list_of_columns <- colnames(df[,grep("fb_metrics.3.values.*", column_names)])

我的问题是我应该如何使用刚刚创建的列名重命名提取的列？还有，有更简单的方法吗？

编辑：

好的，我把它重命名为

library(data.table)
setnames(df, old = list_of_columns, new = sub(pattern = ".*value",replacement = new_column_name,x = colnames(df[,grep(paste("fb_metrics.",i,".values.*",sep = ""), column_names)])))

但有没有更简单的方法来完成整个过程？

Answer 1

当面对复杂/多次替换时，您可以通过创建自定义函数来简化该过程使用不同的输入参数并按顺序应用它以达到所需的列名结构。

<强> InputData：

DF1 = data.frame(fb_metrics.3.name="post_storytellers_by_action_type",fb_metrics.3.period="lifetime",fb_metrics.3.values.0.value.share=1:5,fb_metrics.3.values.0.value.like=1:5,
fb_metrics.3.values.0.value.comment=1:5,stringsAsFactors=FALSE)


DF1
#                 fb_metrics.3.name fb_metrics.3.period fb_metrics.3.values.0.value.share
#1 post_storytellers_by_action_type            lifetime                                 1
#2 post_storytellers_by_action_type            lifetime                                 2
#3 post_storytellers_by_action_type            lifetime                                 3
#4 post_storytellers_by_action_type            lifetime                                 4
#5 post_storytellers_by_action_type            lifetime                                 5

#  fb_metrics.3.values.0.value.like fb_metrics.3.values.0.value.comment
#1                                1                                   1
#2                                2                                   2
#3                                3                                   3
#4                                4                                   4
#5                                5                                   5

<强> CustomFunction：

#Custom function to aid replacements

 fn_modify_str = function(pattern="a",replacement="str",suffix=".",inputString="abcd") {

 gsub(pattern,paste0(replacement,suffix),inputString)

 }

<强> Pararmeters：

 inputColumn1 = "^fb_metrics.3.name$"
 inputColumn2 = "^fb_metrics.3.period$"

 replacePattern1 = "fb_metrics.3.values.0[.]"
 replacePattern2 = "value"

 uniqValue1 = unique(DF1[,grep(inputColumn1,colnames(DF1))])

uniqValue1
#[1] "post_storytellers_by_action_type"

 uniqValue2 = unique(DF1[,grep(inputColumn2,colnames(DF1))])

uniqValue2
#[1] "lifetime"

<强>替换

#apply replacements using custom function  sequentially for both patterns

strPart1 =  fn_modify_str(pattern=replacePattern1,replacement = uniqValue1,suffix=".",inputString = colnames(DF3))

strPart2 =  fn_modify_str(pattern=replacePattern2,replacement = uniqValue2,suffix="",inputString = strPart1)


#you can rename columns in the same dataset by just simple assignment
#colnames(DF1) = strPart2

#OR, you can create a backup dataset and rename the columns in the new DF
DF2 = DF1

colnames(DF2) = strPart2

<强>输出：

DF2
#                 fb_metrics.3.name fb_metrics.3.period post_storytellers_by_action_type.lifetime.share
#1 post_storytellers_by_action_type            lifetime                                               1
#2 post_storytellers_by_action_type            lifetime                                               2
#3 post_storytellers_by_action_type            lifetime                                               3
#4 post_storytellers_by_action_type            lifetime                                               4
#5 post_storytellers_by_action_type            lifetime                                               5
#  post_storytellers_by_action_type.lifetime.like post_storytellers_by_action_type.lifetime.comment
#1                                              1                                                 1
#2                                              2                                                 2
#3                                              3                                                 3
#4                                              4                                                 4
#5                                              5                                                 5

使用另一列中的值重命名列

1 个答案: