Merging based on matching characters in R

Question

For two example dataframes:

df1 <- structure(list(name = c("Katie", "Eve", "James", "Alexander", 
"Mary", "Barrie", "Harry", "Sam"), postcode = c("CB12FR", "CB12FR", 
"NE34TR", "DH34RL", "PE46YH", "IL57DS", "IP43WR", "IL45TR")), .Names = c("name", 
"postcode"), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-8L), spec = structure(list(cols = structure(list(name = structure(list(), class = c("collector_character", 
"collector")), postcode = structure(list(), class = c("collector_character", 
"collector"))), .Names = c("name", "postcode")), default = structure(list(), class = c("collector_guess", 
"collector"))), .Names = c("cols", "default"), class = "col_spec"))

df2 <-structure(list(name = c("Katie", "James", "Alexander", "Lucie", 
"Mary", "Barrie", "Claire", "Harry", "Clare", "Hannah", "Rob", 
"Eve", "Sarah"), postcode = c("CB12FR", "NE34TR", "DH34RL", "DL56TH", 
"PE46YH", "IL57DS", "RE35TP", "IP43WQ", "BH35OP", "CB12FR", "DL56TH", 
"CB12FR", "IL45TR"), rating = c(1L, 1L, 1L, 2L, 3L, 1L, 4L, 2L, 
2L, 3L, 1L, 4L, 2L)), .Names = c("name", "postcode", "rating"
), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-13L), spec = structure(list(cols = structure(list(name = structure(list(), class = c("collector_character", 
"collector")), postcode = structure(list(), class = c("collector_character", 
"collector")), rating = structure(list(), class = c("collector_integer", 
"collector"))), .Names = c("name", "postcode", "rating")), default = structure(list(), class = c("collector_guess", 
"collector"))), .Names = c("cols", "default"), class = "col_spec"))

I want to add an additional column to df1 which gives the ratings from df2. There may be multiple ratings for each postcode (which is why a straight merge wouldn't work.

I only want to merge the two dataframes WHEN the postcode AND the first 3 characters of the name are the same (providing these are unique in df1). for example, if there was a Katherine and Katie - (both with the same postcode), these wouldn't be merged

I am happy to have blanks where there is no merge.

Any ideas?

Answer 1

Wouldn't a simple join with multiple columns solve your problem? Something like,

df<-merge(x=df1,y=df2,by=c('name','postcode'),all.x=T)

Alternate solution in case the column names don't match,

df1$key<-paste(df1$name,df1$postcode,sep="_")
df2$key<-paste(df2$name,df2$postcode,sep="_")

df<-merge(x=df1,y=df2,by=c('key'),all.x=T)