R根據另一列的值添加一列

Question

我正在嘗試轉換具有這樣的數據的data.frame

COMPANY COUNTRY CURRENCY IND1     WGT1 IND2   WGT2 IND3     WGT3
COMP1   USA     USD      HEALTH   .58  RETAIL .42  <NA>     0
COMP2   USA     USD      AUTO     .78  RETAIL .12  TRANSPRT .1
COMP3   CAN     CAD      SOFTWARE 1    <NA>   0    <NA>     0

我想把它變成

COMPANY COUNTRY CURRENCY HEALTH AUTO SOFTWARE RETAIL TRANSPRT
COMP1   USA     USD      .58    0    0        .42    0
COMP2   USA     USD      0      .78  0        .12    .1
COMP3   CAN     CAD      0      0    1        0      0

解決此問題的最佳方法是什么？ 在此先感謝BE

Answer 1

我們可以用melt/dcast從devel版本data.table即v1.9.5 。

將'data.frame'轉換為'data.table'（ setDT(df1) ）。 來自data.table的melt可以采用多個measure列。 我們將以前綴“ IND”，“ WGT”開頭的列名稱指定為patterns ，然后將“ wide”轉換為“ long”格式。 通過將其分配到NULL刪除“變量”列，然后dcast從“長”到“寬”指定“value.var”作為“值”。

library(data.table)#v1.9.5+
DT <- melt(setDT(df1), measure=patterns('^IND', 'WGT'),
                            na.rm=TRUE)[, variable:=NULL]
dcast(DT, ...~value1, value.var='value2', fill=0)
#   COMPANY COUNTRY CURRENCY AUTO HEALTH RETAIL SOFTWARE TRANSPRT
#1:   COMP1     USA      USD 0.00   0.58   0.42        0      0.0
#2:   COMP2     USA      USD 0.78   0.00   0.12        0      0.1
#3:   COMP3     CAN      CAD 0.00   0.00   0.00        1      0.0

數據

df1 <- structure(list(COMPANY = c("COMP1", "COMP2", "COMP3"), 
COUNTRY = c("USA", 
"USA", "CAN"), CURRENCY = c("USD", "USD", "CAD"), IND1 = c("HEALTH", 
"AUTO", "SOFTWARE"), WGT1 = c(0.58, 0.78, 1), IND2 = c("RETAIL", 
"RETAIL", NA), WGT2 = c(0.42, 0.12, 0), IND3 = c(NA, "TRANSPRT", 
NA), WGT3 = c(0, 0.1, 0)), .Names = c("COMPANY", "COUNTRY", "CURRENCY", 
"IND1", "WGT1", "IND2", "WGT2", "IND3", "WGT3"), row.names = c(NA, 
-3L), class = "data.frame")

Answer 2

這是使用reshape和xtabs的默認庫解決方案

long<-reshape(df,sep="",varying=4:9,direction="long")
cbind(df[,1:3],as.data.frame.matrix(xtabs(WGT~COMPANY+IND,long)))

COMPANY COUNTRY CURRENCY AUTO HEALTH SOFTWARE RETAIL TRANSPRT
COMP1   COMP1     USA      USD 0.00   0.58        0   0.42      0.0
COMP2   COMP2     USA      USD 0.78   0.00        0   0.12      0.1
COMP3   COMP3     CAN      CAD 0.00   0.00        1   0.00      0.0