CODEWITHSUNDEEP
rdatetimegroupinglubridate
Tdebeus
Tdebeus

Reputation: 1599

Convert hour:min:sec column of type character to time

I've a time column that's of type character. I want to convert it to some sort of time type. I like to use the lubridate package but I don't like the notation of the output. My goal is to later group the date per hour and count the cases. df %>% group_by(floor_date(Tijd, "hour")).

Dataframe:

df <- structure(list(Tijd = c("01:00:00", "01:27:00", "09:25:00", "14:39:00", 
"15:35:00")), .Names = "Tijd", row.names = c(NA, -5L), class = c("tbl_df", 
"tbl", "data.frame"))

The following code produces a a type where the floor_date() function doesn't work on.

library(tidyverse)
library(lubridate)

df %>% 
  mutate(Tijd = hms(Tijd))

Output:

# A tibble: 5 x 1
  Tijd        
  <S4: Period>
1 1H 0M 0S    
2 1H 27M 0S   
3 9H 25M 0S   
4 14H 39M 0S  
5 15H 35M 0S

The following code gives an error:

df %>%
  mutate(Tijd = hms(Tijd)) %>%
  group_by(floor_date(Tijd, "hour"))

Error in mutate_impl(.data, dots) : 
  Evaluation error: 'origin' must be supplied.

Upvotes: 2

Views: 3048

Answers (2)

Prem
Prem

Reputation: 11955

In case you are interested in chron approach then

library(dplyr)
library(chron)

df %>% 
  mutate(Tijd = hours(times(Tijd)))

gives

   Tijd
  <dbl>
1  1.00
2  1.00
3  9.00
4 14.0 
5 15.0


Sample data:

df <- structure(list(Tijd = c("01:00:00", "01:27:00", "09:25:00", "14:39:00", 
"15:35:00")), .Names = "Tijd", row.names = c(NA, -5L), class = c("tbl_df", 
"tbl", "data.frame"))

Upvotes: 0

Vikash Kumar
Vikash Kumar

Reputation: 348

One of the way -

hr <- strptime(df$Tijd, format = '%H:%M:%S', 'GMT')
df$hr <- hour(hr)

Upvotes: 2

Related Questions