DateTime Diff in R

Question

I need help in creating a function that classifies a record as TRUE if a record is shows a diff of 7 days per User. Note that the User and DateTime fields are not arranged in order, I just arranged it for easier representation of the dataset.

 User     DateTime              Result
 A        2015-05-27 17:13      FALSE
 A        2015-06-23 14:17      FALSE
 A        2015-06-24 15:44      TRUE
 A        2015-06-27 12:16      TRUE
 B        2015-03-04 18:07      FALSE
 C        2015-07-27 08:26      FALSE
 D        2015-03-26 18:13      FALSE
 D        2015-05-20 10:35      FALSE
 D        2015-05-25 18:07      TRUE

Obviously, my function does not work when I tried this because it just gives me one logical value:

 repeatfun <- function(x) {ifelse(sum(diff(x) < 7), TRUE, FALSE)}

Here's the data for easier replication:

User <- c('A', 'A', 'A', 'A', 'A', 'B', 'C', 'D', 'D', 'D', 'D', 'D')
DateTime <- c('2015-05-27', '2015-06-23', '2015-06-24', '2015-06-27', '2015-07-08',
          '2015-03-04', '2015-07-27',
          '2015-03-26', '2015-05-20', '2015-05-25', '2015-06-17', '2015-08-13')
df <- as.data.frame(cbind(User, DateTime))
df$DateTime <- as.Date(df$DateTime)

Pierre L · Accepted Answer · 2015-09-24 15:38:58Z

4

With dplyr, we can group by User and arrange DateTime from the earliest date. Finally to create Result, DateTime is subtracted from the previous date. The argument default=FALSE prevents NA values from appearing. The output is tested with (x < 7).

library(dplyr)
df %>% group_by(User) %>% arrange(DateTime) %>% 
  mutate(Result=DateTime-lag(DateTime, default=F) < 7)
# Source: local data frame [9 x 3]
# Groups: User [4]
# 
#     User   DateTime Result
#   (fctr)     (date)  (lgl)
# 1      A 2015-05-27  FALSE
# 2      A 2015-06-23  FALSE
# 3      A 2015-06-24   TRUE
# 4      A 2015-06-27   TRUE
# 5      B 2015-03-04  FALSE
# 6      C 2015-07-27  FALSE
# 7      D 2015-03-26  FALSE
# 8      D 2015-05-20  FALSE
# 9      D 2015-05-25   TRUE

answered Sep 24, 2015 at 15:38

Pierre L

28.5k6 gold badges50 silver badges70 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

LyzandeR · Accepted Answer · 2015-09-24 15:50:53Z

3

Another solution using data.table and your slightly changed function:

#your function without the sum function
repeatfun <- function(x) {ifelse(diff(x) < 7, TRUE, FALSE)}

#data.table solution
setDT(df)[, ,key='DateTime'][, Result := c(FALSE, repeatfun(DateTime)), by=User]

Output:

> df
   User   DateTime Result
1:    A 2015-05-27  FALSE
2:    A 2015-06-23  FALSE
3:    A 2015-06-24   TRUE
4:    A 2015-06-27   TRUE
5:    B 2015-03-04  FALSE
6:    C 2015-07-27  FALSE
7:    D 2015-03-26  FALSE
8:    D 2015-05-20  FALSE
9:    D 2015-05-25   TRUE

answered Sep 24, 2015 at 15:50

LyzandeR

38k12 gold badges84 silver badges90 bronze badges

Collectives™ on Stack Overflow

DateTime Diff in R

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related