How to select only the last row among the subset of rows satisfying a condition in R programming

Question

The dataframe looks like this :

Customer_id A B C D E F G
10000001    1 1 2 3 1 3 1
10000001    1 2 3 1 2 1 3
10000002    2 2 2 3 1 3 1
10000002    2 2 1 4 2 3 1         


        

Answer 1

Something like this (hard to code without the data in R format):

dataframe[ rev(!duplicated(rev(dataframe$Customer_id))),]

or better

dataframe[ !duplicated(dataframe$Customer_id,fromLast=TRUE),]

Answer 2

You can also use aggregate

aggregate(. ~ Customer_id, data = DF, FUN =  tail, 1)
##   Customer_id A B C D E F G
## 1    10000001 1 2 3 1 2 1 3
## 2    10000002 2 2 1 4 2 3 1
## 3    10000003 1 1 2 2 1 2 1
## 4    10000004 1 4 1 4 1 3 1
## 5    10000006 1 3 1 4 1 2 1
## 6    10000008 1 3 1 1 2 2 1

Answer 3

Assume your data is named dat,

Here's one way using by and rbind, although the other two methods (aggregate and duplicated) are much nicer:

> do.call(rbind, by(dat,dat$Customer_id,FUN=tail,1))
##    Customer_id A B C D E F G
## 2     10000001 1 2 3 1 2 1 3
## 4     10000002 2 2 1 4 2 3 1
## 7     10000003 1 1 2 2 1 2 1
## 11    10000004 1 4 1 4 1 3 1
## 13    10000006 1 3 1 4 1 2 1
## 16    10000008 1 3 1 1 2 2 1