如何停止 tidyr 按字母顺序传播排序列

Question

我在“SampleFileName”列中有一个带有随机顺序的样本名称的表“inputdf”。

> colnames(inputdf)
 [1] "Dye/SamplePeak" "SampleFileName" "Marker"         "Allele"         "Size"           "Height"        
 [7] "Area"           "DataPoint"      "flank"          "correction"     "start"          "end"           
[13] "control"        "iithreshold"    "CAG"           

我使用 tidyr 将“高度”列中的结果分散到单独的列中，每列由“SampleFileName”中的值命名。

library(tidyr)
height <- spread(inputdf, key=SampleFileName, value=Height, fill = 0, convert = FALSE) #Extract heights into separate columns for each sample

我的样本在“SampleFileName”列中没有按字母顺序排列，我想保持这种顺序。但是，spread 会自动按字母顺序对它们进行排序。非常感谢您的帮助！

> colnames(height)
 [1] "Dye/SamplePeak"                         "Marker"                                
 [3] "Allele"                                 "Size"                                  
 [5] "Area"                                   "DataPoint"                             
 [7] "flank"                                  "correction"                            
 [9] "start"                                  "end"                                   
[11] "control"                                "iithreshold"                           
[13] "CAG"                                    "A01_MF20170522_FA_A01_2017-05-22_1.fsa"
[15] "A01_MF20170623_FA_A01_2017-06-23_1.fsa" "A02_MF20170623_FA_A02_2017-06-23_1.fsa"
[17] "A03_MF20170623_FA_A03_2017-06-23_1.fsa" "A05_MF20170623_FA_A05_2017-06-23_1.fsa"
[19] "A06_MF20170623_FA_A06_2017-06-23_1.fsa" "A07_MF20170623_FA_A07_2017-06-23_1.fsa"
[21] "A08_MF20170623_FA_A08_2017-06-23_1.fsa" "A09_MF20170623_FA_A09_2017-06-23_1.fsa"
[23] "A10_MF20170623_FA_A10_2017-06-23_1.fsa" "A11_MF20170623_FA_A11_2017-06-23_1.fsa"
[25] "A12_MF20170623_FA_A12_2017-06-23_1.fsa" "B01_MF20170623_FA_B01_2017-06-23_1.fsa"
[27] "B02_MF20170623_FA_B02_2017-06-23_1.fsa" "B03_MF20170623_FA_B03_2017-06-23_1.fsa"
[29] "B04_MF20170623_FA_B04_2017-06-23_1.fsa" "B05_MF20170623_FA_B05_2017-06-23_1.fsa"
[31] "B06_MF20170623_FA_B06_2017-06-23_1.fsa" "B07_MF20170623_FA_B07_2017-06-23_1.fsa"
[33] "B08_MF20170522_FA_B08_2017-05-22_1.fsa" "B08_MF20170623_FA_B08_2017-06-23_1.fsa"
[35] "C01_MF20170623_FA_C01_2017-06-23_1.fsa" "C02_MF20170529_FA_C02_2017-05-30_1.fsa"
[37] "C02_MF20170623_FA_C02_2017-06-23_1.fsa" "C05_MF20170623_FA_C05_2017-06-23_1.fsa"
[39] "C07_MF20170623_FA_C07_2017-06-23_1.fsa" "C08_MF20170623_FA_C08_2017-06-23_1.fsa"
[41] "C09_MF20170623_FA_C09_2017-06-23_1.fsa" "C10_MF20170623_FA_C10_2017-06-23_1.fsa"
[43] "C11_MF20170623_FA_C11_2017-06-23_1.fsa" "C12_MF20170623_FA_C12_2017-06-23_1.fsa"
[45] "D02_MF20170623_FA_D02_2017-06-23_1.fsa" "D03_MF20170623_FA_D03_2017-06-23_1.fsa"
[47] "D04_MF20170623_FA_D04_2017-06-23_1.fsa" "D05_MF20170623_FA_D05_2017-06-23_1.fsa"
[49] "D06_MF20170623_FA_D06_2017-06-23_1.fsa" "D08_MF20170623_FA_D08_2017-06-23_1.fsa"
[51] "D10_MF20170623_FA_D10_2017-06-23_1.fsa" "D11_MF20170623_FA_D11_2017-06-23_1.fsa"
[53] "D12_MF20170623_FA_D12_2017-06-23_1.fsa"

Answer 1

您可以在tidyr::spread() 之后使用dplyr::select() 来保留原始列顺序。请注意，您的行可能仍按字母顺序排序。

library(tidyverse)

mtcars <- rownames_to_column(mtcars, "vehicle")

colnames(mtcars)
# [1] "vehicle" "mpg"     "cyl"     "disp"    "hp"      "drat"    "wt"      "qsec"    "vs"      "am"      "gear"    "carb"

按字母顺序排列的列顺序的常规传播行为：

spreadMtcarsSorted <- gather(mtcars, measure, value, -vehicle) %>%
  spread(measure, value)

colnames(spreadMtcarsSorted)
# [1] "vehicle" "am"      "carb"    "cyl"     "disp"    "drat"    "gear"    "hp"      "mpg"     "qsec"    "vs"      "wt" 

展开操作后恢复到原来的列顺序：

spreadMtcarsOriginalColOrder <- gather(mtcars, measure, value, -vehicle) %>%
  spread(measure, value) %>%
  select(colnames(mtcars))  # use original column order before returning new data frame

colnames(spreadMtcarsOriginalColOrder)
# [1] "vehicle" "mpg"     "cyl"     "disp"    "hp"      "drat"    "wt"      "qsec"    "vs"      "am"      "gear"    "carb"

Answer 2

你可以试试这样的。

library(tidyr)

# Get vector of current column names (excluding "SampleFileName" and "Height" as they will not exist in final dataset) and all of the SampleFileName values.
cols <- c(colnames(inputdf)[!(colnames(inputdf) %in% c("SampleFileName","Height"))], unique(inputdf$SampleFileName))

# Spread the SampleFileName column
height <- spread(inputdf, key=SampleFileName, value=Height, fill = 0, convert = FALSE) 

# Select the columns in the order they are listed in the cols vector
height <- height[,cols]