按多个条件匹配和替换数据框的列

Question

干杯,我有两个数据框,具有以下结构.

DF1:
Airlines           HeadQ      Date           Cost_Index
American           PHX        07-31-2016     220
American           ATL        08-31-2016     150
American           ATL        10-31-2016     150
Delta              ATL        10-31-2016     180
American           ATL        08-31-2017     200

第二数据帧DF2具有以下结构:

DF2:
Airlines           HeadQ      Date          
American           ATL        09-30-2016
Delta              ATL        03-31-2017

现在查看数据帧DF1和DF2,我想将DF1改为以下数据帧.

DF1:
Airlines           HeadQ      Date           Cost_Index
American           PHX        07-31-2016     220
American           ATL        08-31-2016     0
American           ATL        10-31-2016     150
Delta              ATL        10-31-2016     180
American           ATL        08-31-2017     200

条件是,从DF2查找DF1的航空公司和HeadQ,如果DF1 $ Date <DF2 $ Date,则将Cost_Index设为0,否则继续使用Cost_Index.

我尝试了,但未成功,:

DF1$Cost_Index <- ifelse(DF1$Airlines == DF2$Airlines & DF1$HeadQ == DF2$HeadQ 
        & DF1$Date < DF2$Date, 0, DF1$Cost_Index)


Warning:
1: In DF1$Airlines == DF2$Airlines : longer object
length is not a multiple of shorter object length". 
2: In<=.default(DF1$Date, DF2$Date) : longer object length is not a
multiple of shorter object length

DF1:
Airlines           HeadQ      Date           Cost_Index
American           PHX        07-31-2016     220
American           ATL        08-31-2016     0
American           ATL        10-31-2016     0
Delta              ATL        10-31-2016     0
American           ATL        08-31-2017     200

有人能指出我正确的方向吗？

注意:

str(DF1$Date): Date, format: "2016-10-31"
str(DF2$Date): Date, format: "2016-08-31"

Answer 1

使用 条件连接功能(自1.9.8),我将按如下方式执行此操作:

require(data.table) # v1.9.8+
# convert to data.tables, and Date column to Date class.
setDT(df1)[, Date := as.Date(Date, format = "%m-%d-%Y")]
setDT(df2)[, Date := as.Date(Date, format = "%m-%d-%Y")]

df1[df2, on = .(Airlines, HeadQ, Date < Date), # find matching rows based on condition
      Cost_Index := 0L]                        # update column with 0 for those rows

df1
#    Airlines HeadQ       Date Cost_Index
# 1: American   PHX 2016-07-31        220
# 2: American   ATL 2016-08-31          0
# 3: American   ATL 2016-10-31        150
# 4:    Delta   ATL 2016-10-31        180