如何在R中將數據從一行分為兩行？

Question

對於在“ Submit.and.module”列中具有“ Separate”值的行，我想在其上方直接插入一行並將數據從某些列移至該新行。

具體來說，我想將名為“ Submit.help”和“策略”的列中的數據移動到上面的新行。

現在，我的數據如下所示：

我希望數據看起來像這樣：

我怎樣才能做到這一點？

Answer 1

以為我會加入另一種解決方案參加聚會。

數據

data <- data.frame(Which.mod         = c("TMH", "TMH-C", "TMH", "FC", "FC"),
                   Mod.time          = c(1.43, 2.31, 0.67, 2.35, 8.22),
                   Submit.help       = c(NA, "Help", NA, NA, "Submit"),
                   Strategy          = c(NA, "Ratio", NA, NA, "Count"),
                   Submit.and.module = c(NA, "Separate", NA, NA, "Separate"))

基數R

步驟1：制作一個ID列和一個新的數據框，該框是要分離的行的子集。

data$id <- 1:nrow(data)
data1 <- subset(data, !is.na(Submit.and.module))

步驟2：將適當的列設置為NA並綁定數據幀

data[, c("Submit.help", "Strategy")] <- NA
data1[, c("Which.mod", "Mod.time", "Submit.and.module")] <- NA

步驟3：綁定數據幀和順序。

final <- rbind(data1, data)
final.ordered <- df1[order(df1$id), ]

#    Which.mod Mod.time Submit.help Strategy Submit.and.module id
# 1        TMH     1.43        <NA>     <NA>              <NA>  1
# 2       <NA>       NA        Help    Ratio              <NA>  2
# 21     TMH-C     2.31        <NA>     <NA>          Separate  2
# 3        TMH     0.67        <NA>     <NA>              <NA>  3
# 4         FC     2.35        <NA>     <NA>              <NA>  4
# 5       <NA>       NA      Submit    Count              <NA>  5
# 51        FC     8.22        <NA>     <NA>          Separate  5

Tidyverse

尼斯，易於遵循。 與上述相同的步驟，但是要盡可能多地鏈接。

library(tidyverse)
dat1 <- data %>% mutate(id = 1:n(), Submit.help = NA, Strategy = NA)
dat2 <- data %>% mutate(id = 1:n()) %>% 
                 filter(!is.na(Submit.and.module)) %>%
                 mutate(Which.mod = NA, Mod.time = NA, Submit.and.module = NA)
final <- rbind(dat2, dat1) %>% arrange(id)

Answer 2

這個想法是將要復制的列剪切到一個單獨的數據表中，將其添加到原始數據中（最后），並使用名為“ sort”的幫助器列以正確的順序對行進行排序：

library(data.table)

data <- data.table(Which.mod         = c("TMH", "TMH-C", "TMH", "FC", "FC"),
                   Mod.time          = c(1.43, 2.31, 0.67, 2.35, 8.22),
                   Submit.help       = c(NA, "Help", NA, NA, "Submit"),
                   Strategy          = c(NA, "Ratio", NA, NA, "Count"),
                   Submit.and.module = c(NA, "Separate", NA, NA, "Separate"))


data[, sort := (1:.N) * 10]  # add a row sorting value with a gap to fill in inserted rows

# cut columns to be duplicated and "insert" at the end    
res <- rbind(data,
             data[Submit.and.module == "Separate", .(sort, Submit.help, Strategy)] [, sort := sort - 1],
             use.names = TRUE,
             fill = TRUE)

# Purge content of moved columns (credits go to @ismiregal - I forgot this initially)
res[Submit.and.module %in% "Separate", c("Submit.help", "Strategy") := NA]

# sort the result accordingly
res <- res[order(sort),]

結果：

res

   Which.mod Mod.time Submit.help Strategy Submit.and.module sort
1:       TMH     1.43        <NA>     <NA>              <NA>   10
2:      <NA>       NA        Help    Ratio              <NA>   19
3:     TMH-C     2.31        <NA>     <NA>          Separate   20
4:       TMH     0.67        <NA>     <NA>              <NA>   30
5:        FC     2.35        <NA>     <NA>              <NA>   40
6:      <NA>       NA      Submit    Count              <NA>   49
7:        FC     8.22        <NA>     <NA>          Separate   50

Answer 3

代碼很丑陋，但是如果您需要的話，我會盡力解釋它的作用（如果我早上會記得的話）。 當然，還有另一種優雅的方法可以做到這一點。 但是只是為了慶祝多樣性...

您的示例數據集：

dat <- data.frame(
  Wich.mod = c("TMH", "TMH-C", "TMH", "FC", "FC"),
  Mod.time = c(1.43, 2.31, 0.67, 2.35, 8.22),
  Submit.help = c(NA, "Help", NA, NA, "Submit"),
  Strategy = c(NA, "Ratio", NA, NA, "Count"),
  Submit.and.module = c(NA, "Separate", NA, NA, "Separate"),
  stringsAsFactors = F
)

Transfromations：

## create new data.frame, filled with NA. The same cols, but extra N("Separate") rows

newdata <- data.frame(matrix(NA, nrow(dat), ncol(dat) + sum(grepl("Separate", dat[, 5]))))

## insert data from dat, leaving empty spaces before "Separate"

newdata[1:nrow(dat) + cumsum(grepl("Separate", dat[, 5])), ] <- dat[1:nrow(dat),]

## give newdata column names from old data

colnames(newdata) <- colnames(dat)

## move Submit.help and Strategy related to "Separate" a row up

newdata[
  which(newdata[, 5] == "Separate") - 1, 3:4
  ] <- newdata[which(newdata[, 5] == "Separate"), 3:4]

## for variables above, replace old values related to "Separate" with NA

newdata[which(newdata[, 5] == "Separate"), 3:4] <- NA

輸出：

#  Wich.mod Mod.time Submit.help Strategy Submit.and.module
# 1 TMH          1.43 NA          NA       NA               
# 2 NA           NA   Help        Ratio    NA               
# 3 TMH-C        2.31 NA          NA       Separate         
# 4 TMH          0.67 NA          NA       NA               
# 5 FC           2.35 NA          NA       NA               
# 6 NA           NA  Submit       Count    NA               
# 7 FC           8.22 NA          NA       Separate  

Answer 4

這是另一個data.table解決方案（看來我太慢了）：

library(data.table)

DT <- data.table(stringsAsFactors=FALSE,
               Index = seq(5),
           Which.mod = c("TMH", "TMH-C", "TMH", "FC", "FC"),
            Mod.time = c(1.43, 2.31, 0.67, 2.35, 8.22),
         Submit.help = c(NA, "Help", NA, NA, "Submit"),
            Strategy = c(NA, "Ratio", NA, NA, "Count"),
   Submit.and.module = c(NA, "Separate", NA, NA, "Separate")
)

DT <- rbindlist(list(DT, DT[Submit.and.module %in% "Separate", c("Index", "Submit.help", "Strategy")]), use.names=TRUE, fill=TRUE)
DT[Submit.and.module %in% "Separate", c("Submit.help", "Strategy") := NA]
setorder(DT, Index, Mod.time, na.last=FALSE)
print(DT)

Answer 5

這是一個tidyverse解決方案：

library(tidyverse)
nms <- names(df1)
df1 %>% 
  rowid_to_column %>%
  gather(,,-rowid) %>% 
  filter(!is.na(value)) %>%
  mutate(tmp = key %in% c("Which.mod","Mod.time","Submit.and.module")) %>%
  spread(key,value) %>%
  select_at(nms)

#   Which.mod Mod.time Submit.help Strategy Submit.and.module
# 1       TMH     1.43        <NA>     <NA>              <NA>
# 2      <NA>     <NA>        Help    Ratio              <NA>
# 3     TMH-C     2.31        <NA>     <NA>          Separate
# 4       TMH     0.67        <NA>     <NA>              <NA>
# 5        FC     2.35        <NA>     <NA>              <NA>
# 6      <NA>     <NA>      Submit    Count              <NA>
# 7        FC     8.22        <NA>     <NA>          Separate

數據

df1 <- data.frame(
  Which.mod = c("TMH", "TMH-C", "TMH", "FC", "FC"),
  Mod.time = c(1.43, 2.31, 0.67, 2.35, 8.22),
  Submit.help = c(NA, "Help", NA, NA, "Submit"),
  Strategy = c(NA, "Ratio", NA, NA, "Count"),
  Submit.and.module = c(NA, "Separate", NA, NA, "Separate"),
  stringsAsFactors = FALSE
)