重新排序名称中包含日期的数据框列？

Question

我有一个反应性数据框，其中列名发生变化，Month.Year 名称的列出现故障。我怎样才能把第一个 Month.Year 放在“Current”之后最左边的位置？以下是数据框列的排序方式以及我希望它们的排列方式。

print(colnames(df))
#[1] "ProductCategoryDesc" "RegionDesc"          "SourceDesc"          "Report"             
#[5] "Apr.2019"            "Current"             "Feb.2019"            "Jun.2019"           
#[9] "Mar.2019"            "May.2019"            "Mar.2020"

#the order I want is below
#[1] "ProductCategoryDesc" "RegionDesc"          "SourceDesc"          "Report"             
#[5] "Current"             "Feb.2019"             "Mar.2019"            "Jun.2019"           
#[9] "Apr.2019"            "May.2019"             "Mar.2020"

#####################################################################
#another example of the df
print(colnames(df))

#[1] "ProductCategoryDesc" "RegionDesc"          "SourceDesc"          "Report"             
#[5] "Apr.2019"            "Current"             "Feb.2019"            "Jun.2019"           
#[9] "Mar.2019"            "May.2019"            "Sep.2019"

#the order I want is below
#[1] "ProductCategoryDesc" "RegionDesc"          "SourceDesc"          "Report"             
#[5] "Current"             "Feb.2019"             "Mar.2019"            "Apr.2019"           
#[9] "May.2019"            "Jun.2019"             "Sep.2019"

这里是一些关于 df 外观的信息

print(dput(droplevels(head(d3))))
#below is the output

structure(list(ProductCategoryDesc = structure(c(1L, 1L, 1L, 
1L, 1L, 1L), .Label = "CN AMMONIA", class = "factor"), RegionDesc = 
structure(c(1L, 
1L, 1L, 1L, 1L, 1L), .Label = "AB REG 2 UPPER MIDWEST", class = "factor"), 
SourceDesc = structure(c(1L, 1L, 1L, 1L, 1L, 1L), .Label = "CN-SD, WATERTOWN 
LIQUID", class = "factor"), 
Report = structure(1:6, .Label = c("InventoryAvailabletoShip", 
"NetCashPosition", "NetMarketPositionTotal", "NonDirectShipPurchase", 
"TotalDirectShips", "TotalNonDirectShips"), class = "factor"), 
Apr.2019 = c(0, 0, 0, 0, 0, 0), Current = c(0, 0, 0, 0, 0, 
0), Feb.2019 = c(0, 0, 240, 240, 0, 240), Jun.2019 = c(0, 
0, 0, 0, 0, 0), Mar.2019 = c(0, 0, 0, 0, 0, 0), May.2019 = c(0, 
0, 0, 0, 0, 0)), sorted = c("ProductCategoryDesc", "RegionDesc", 
"SourceDesc", "Report"), row.names = c(NA, -6L), .internal.selfref = 
<pointer: 0x0000000000211ef0>, class = c("data.table", 
"data.frame"))
ProductCategoryDesc             RegionDesc              SourceDesc                   
Report Apr.2019
1:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID 
InventoryAvailabletoShip        0
2:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID          
NetCashPosition        0
3:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID   
NetMarketPositionTotal        0
4:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID    
NonDirectShipPurchase        0
5:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID         
TotalDirectShips        0
6:          CN AMMONIA AB REG 2 UPPER MIDWEST CN-SD, WATERTOWN LIQUID      
TotalNonDirectShips        0
Current Feb.2019 Jun.2019 Mar.2019 May.2019
1:       0        0        0        0        0
2:       0        0        0        0        0
3:       0      240        0        0        0
4:       0      240        0        0        0
5:       0        0        0        0        0
6:       0      240        0        0        0

Answer 1

我们可以尽可能转换为日期，并对列进行排序：

x <- c("ProductCategoryDesc", "RegionDesc","SourceDesc","Report",             
 "Apr.2019","Current","Feb.2019", "Jun.2019",           
 "Mar.2019","May.2019","Mar.2020")

dates <-  as.Date(paste0("01.",x), "%d.%b.%Y")
x <- x[order(replace(dates, is.na(dates), "0000-01-01"))]
# [1] "ProductCategoryDesc" "RegionDesc"          "SourceDesc"          "Report"             
# [5] "Current"             "Feb.2019"            "Mar.2019"            "Apr.2019"           
# [9] "May.2019"            "Jun.2019"            "Mar.2020"         

您的排序数据框：

df[x]

Answer 2

享受吧！

# Reorder columns in dataframe
df =
  df[
    c("ProductCategoryDesc",
      "RegionDesc", 
      "SourceDesc",
      "Report",
      "Current",
      "Feb.2019",
      "Mar.2019",
      "Jun.2019",
      "Apr.2019",
      "May.2019",
      "Mar.2020")]

要改为按位置（即索引）更改数据帧的顺序，请尝试：

df = 
  df[c(
       1,
       3,
       2)]

如果您能够预测每次将数据添加到数据帧时数据帧的哪个索引将是最新的，您可以编写脚本以获取该数据帧并将其移动到您想要的位置。例如，如果它是数据框中的最后一列，并且您想开始将其移动到第一个位置，则可以尝试使用 length(df) 之类的方法并使用长度来指示最后一个元素：

df = 
  df[c(
       length(df),
       1,
       3,
       2)]