Sort table unique parameters

Question

I have a table called myTable (input) :

 user_name   session_num  
1     "Joe"            1    
2     "Tom"            2    
3    "Fred"            1    
4     "Tom"            1    
5     "Joe"            2    
6     "John"           1 

I want to know how many of my user_id have only session_num = 1 (output) :

   user_name   session_num   
1     "Fred"             1
2     "John"             1

Answer 1

Here's a possible solution using data.table

library(data.table)
setDT(df)[, if(all(session_num == 1)) .SD, by = user_name]
#    user_name session_num
# 1:      Fred           1
# 2:      John           1

Another option is to try an anti join

df[session_num == 1][!df[session_num != 1], on = "user_name"]
#    user_name session_num
# 1:      Fred           1
# 2:      John           1

Answer 2

A comparable solution with dplyr:

library(dplyr)
myTable %>%
  group_by(user_name) %>%
  filter(all(session_num == 1))

which gives:

  user_name session_num
     (fctr)       (int)
1      Fred           1
2      John           1

Answer 3

Alternatively we could simply exclude all users that have a session number other than 1, using base R.

# User's with session number other than 1
two <- myTable$user_name[myTable$session_num != 1] 

# Exclude them
myTable[!myTable$user_name %in% two,]
#  user_name session_num
#3      Fred           1
#6      John           1

Answer 4

This is a 2 line answer:

library(data.table)
data1<-fread("test.csv")
data1[user_name == names(which(table(data1$user_name)==1)),][session_num==1,]

First it goes and looks at who is in the dataset only once and then secondly subsets on those where session_num==1.