How to combine R dataframes based constraints on a time column

Question

I have two R tables, each with a list of users and a timestamp corresponding to the time that they took a certain action.

The first of these (df1) two tabl

Answer 1

Part 1 - Original Question

The first part of your question can be answered with the sqldf package.

library(sqldf)
df3 <- sqldf("SELECT * FROM df1 a 
             LEFT JOIN df2 b ON a.time < b.time 
             AND a.user = b.user")[,c(1:2, 4)]

#rename to match OP post
names(df3) <- c("user", "time_1", "time_2")

> df3
  user              time_1              time_2
1    1 2016-12-01 08:53:20 2016-12-01 11:50:11
2    1 2016-12-01 12:45:47                
3    2 2016-12-01 15:34:54                
4    3 2016-12-01 00:49:50 2016-12-01 01:19:10

Part 2 - Time Window

If you want a window of time to allow for the match, you can subtract seconds within the SQL statement as follows:

df3 <- sqldf("SELECT * FROM df1 a 
             LEFT JOIN df2 b ON a.time < (b.time - 10000)
             AND a.user = b.user")[,c(1:2, 4)]
> df3
  user                time              time.1
1    1 2016-12-01 08:53:20 2016-12-01 11:50:11
2    1 2016-12-01 12:45:47                
3    2 2016-12-01 15:34:54                
4    3 2016-12-01 00:49:50                

Note, whatever you select from b.time will be in seconds.