R: find all observations from data.table with values from another data.table

Question

I have two data tables in R: DT1 having about 30K obervations of 5 variables: userID , userName , productID , productName , usersRate , DT2 having only 500 observations of 2 variables: productID , similarProductID . I want to find all rows from DT1 which productID is same as similarProductID from DT2. I've tried DT1[which(DT1$productID==DT2$similarProductID)] and DT1[which(DT1$productID==intersect(DT1$productID,DT2$similarProductID))] but it didn't worked out, I receive too few observations. Any ideas how could I do this?

Answer 1

Fastest way is with a join:

#mock data
DT1<-data.table(userID=1:30000,userName=sample(LETTERS,30000,T),productID=30001:60000,productName=sample(LETTERS,30000,T),userRate=runif(30000))
DT2<-data.table(productID=1:500,similarProductId=sample(30001:60000,500))

#set keys
setkey(DT1,productID)
setkey(DT2,similarProductId)

#join
DT1[DT2]

   productID userID userName productName   userRate productID.1
1:     30014     14        L           R 0.87649196         473
2:     30025     25        E           A 0.02237395         326
3:     30027     27        H           Z 0.43986360         198
4:     30065     65        V           K 0.33047666         240
5:     30123    123        R           X 0.38637559         210
---                                                             
  496:     59575  29575        U           A 0.41036652         214
497:     59665  29665        C           E 0.67345907          45
498:     59724  29724        F           Y 0.18853101          81
499:     59764  29764        D           X 0.50271854         386
500:     59790  29790        Z           A 0.02222698         397

Answer 2

I doubt that this way would be faster than Troy's join answer, but you can use the merge function in R.

  DT1<-data.frame(userID=1:30000,userName=sample(LETTERS,30000,T),productID=30001:60000,productName=sample(LETTERS,30000,T),userRate=runif(30000))
  DT2<-data.frame(productID=sample(30001:60000,500), similarproductID=sample(30001:60000,500))

  colnames(DT2)<-c("BadproductID","productID") #Do this to match the colname in DT1

  DTMerged<-merge(DT1,DT2, by="productID") #Should give you all your matches without NA's