使用合并或匹配函数匹配R中两个数据帧中的多个列

Question

I have two data frames that look something like this:

Date        Shop    Item    ProductKey     Price
2014-09-01  Asda    Apple   0f-7c-32-9c65  2.00
2014-09-01  Tesco   Pear    7c-e9-a0-a11c  1.50

And so on for, for different dates, shops, items, product keys and prices. Data frame two is of the same structure but for the following year.

What I want to do is match items of the same date, shop, item and product key in the two different data frames (call them September2014 and September2015) - and when they match on all variables to create a price relative (ie divide the 2015 price by the 2014 price).

I have tried various if statements and the match function but don't seem to be getting anywhere. I know there must be a simple way to do this that I am completely missing. Any help would be greatly appreciated. I have also looked at examples of the merge function but I don't think that would be useful in my case. I have gone through many questions on the site involving matching and attempted to use some suggested code, but again none seem to be relevant in my case.

Answer 1

Reconsider the merge approach:

# FIRST DATAFRAME (2014)
txt='Date        Shop    Item    ProductKey     Price
2014-09-01  Asda    Apple   0f-7c-32-9c65  2.00
2014-09-01  Tesco   Pear    7c-e9-a0-a11c  1.50'

df1 <- read.table(text=txt, header=TRUE)
df1$Date <- as.POSIXct(df1$Date)             # CONVERT TO DATE
df1$Month <- format(df1$Date, "%m")          # EXTRACT MONTH (CAN ADJUST FOR MM/DD)

# SECOND DATAFRAME (2015)
txt='Date        Shop    Item    ProductKey     Price
2015-09-01  Asda    Apple   0f-7c-32-9c65  2.25
2015-09-01  Tesco   Pear    7c-e9-a0-a11c  1.75'

df2 <- read.table(text=txt, header=TRUE)
df2$Date <- as.POSIXct(df2$Date)              # CONVERT TO DATE
df2$Month <- format(df2$Date, "%m")           # EXTRACT MONTH (CAN ADJUST FOR MM/DD)

# MERGE AND TRANSFORM FOR NEW COLUMN
finaldf <- transform(merge(df1, df2, by=c("Month", "Shop", "Item", "ProductKey"), suffixes=c("_14", "_15")), 
                     PriceRelative = Price_15 / Price_14)    
finaldf
#   Month  Shop  Item    ProductKey    Date_14 Price_14    Date_15 Price_15 PriceRelative
# 1    09  Asda Apple 0f-7c-32-9c65 2014-09-01      2.0 2015-09-01     2.25      1.125000
# 2    09 Tesco  Pear 7c-e9-a0-a11c 2014-09-01      1.5 2015-09-01     1.75      1.166667

Answer 2

Prices in 2014 and 2015. Note that in 2015 there is an item that is not matched by one in 2014. First generate a hash as a key and then match to import the 2014 price of the item into your 2015 dataframe. Then Divide:

df2014 <- data.frame(Date = as.Date(c("2014-09-01", "2014-09-01")),
                     Shop = c("Asda", "Tesco"),
                     Item = c("Apple", "Pear"),
                     ProductKey = c("0f-7c-32-9c65","7c-e9-a0-a11c"),
                     Price = c(2.00, 1.50), stringsAsFactors = FALSE)

df2015 <- data.frame(Date = as.Date(c("2015-09-01", "2015-09-01", "2015-09-01")),
                     Shop = c("Asda", "Tesco", "foo"),
                     Item = c("Apple", "Pear", "Orange"),
                     ProductKey = c("0f-7c-32-9c65","7c-e9-a0-a11c", "blah"),
                     Price = c(2.20, 1.70, 3.00), stringsAsFactors = FALSE)

df2014$key <- paste0(strftime(df2014$Date, "%m"),
                     strftime(df2014$Date, "%d"),
                     df2014$Shop,
                     df2014$Item,
                     df2014$ProductKey)

df2015$key <- paste0(strftime(df2015$Date, "%m"),
                     strftime(df2015$Date, "%d"),
                     df2015$Shop,
                     df2015$Item,
                     df2015$ProductKey)

df2015$price_2014 <- df2014$Price[match(df2015$key, df2014$key)]
df2015$price_ratio <- df2015$Price/df2015$price_2014
df2015