CODEWITHSUNDEEP
rdatestrptime
Mpizos Dimitris
Mpizos Dimitris

Reputation: 4991

Converting Text to specific date format in R

I have the following characters(sample):

>trial
[1] "9-5"  "7-5"  "13-3" "17-6"

And I want to convert them to date format(if possible) and become something like the following:

>trial
[1] Friday 9:00  Friday 7:00  Wednesday 13:00  Saturday 17:00

I am using the following code, which is not the desired result:

>strptime(trial,format="%k-%w")
[1] "2015-10-06 09:00:00 CEST" "2015-10-06 07:00:00 CEST"
[3] "2015-10-06 13:00:00 CEST" "2015-10-06 17:00:00 CEST"

My point is that I want to have date format for weekdays and hour. How can it be done?

Upvotes: 0

Views: 115

Answers (2)

Pierre L
Pierre L

Reputation: 28441

Try:

paste(weekdays((as.Date("1970-01-01")+4)+0:7)[as.numeric(sub(".*-(\\d+)", "\\1", trial))], sub("(\\d+).*", "\\1:00", trial))
[1] "Friday 9:00"     "Friday 7:00"     "Wednesday 13:00"
[4] "Saturday 17:00"

Or to explain what is going on I'll show it in intermediate steps:

#1. starting vector
trial <- c("9-5",  "7-5",  "13-3", "17-6")

#2. get the hour and add ":00" to the values
hr <- sub("(\\d+).*", "\\1:00", trial)

#3. get the day of the week as a number
d <- as.numeric(sub(".*-(\\d+)", "\\1", trial))

#4. pick a week with a Monday date that you know and add 6 days (I use The first Monday after the origin). Subset with 'd'
day_vector <- weekdays((as.Date("1970-01-01")+4)+0:7)[d]

#5. paste day and hour together
paste(day_vector, hr)

[1] "Friday 9:00"     "Friday 7:00"     "Wednesday 13:00"
[4] "Saturday 17:00"

Upvotes: 1

Frank
Frank

Reputation: 66819

Here's what I might do:

library(data.table) # for wday & tstrsplit

a_wk  <- Sys.Date()+1:7
wdays <- weekdays(a_wk)[ order(wday(a_wk)) ]
# "Sunday"    "Monday"    "Tuesday"   "Wednesday" "Thursday"  "Friday"    "Saturday"

data.table(trial)[,tstrsplit(trial, "-", type.convert=TRUE)][, 
  paste0(wdays[V2 + 1], " ", V1, ":00")
]
# "Friday 9:00"     "Friday 7:00"     "Wednesday 13:00" "Saturday 17:00"

The + 1 is needed because the wday function takes Sunday as day 1; while the OP uses Monday.

Comments. It could be written more concisely (only invoking trial once) as

setDT(tstrsplit(trial, "-", type.convert=TRUE))[, 
  paste0(wdays[V2 + 1], " ", V1, ":00")
]

You can see how this works by having a look at the intermediate result:

data.table(trial)[,tstrsplit(trial, "-", type.convert=TRUE)]
#    V1 V2
# 1:  9  5
# 2:  7  5
# 3: 13  3
# 4: 17  6

Upvotes: 2

Related Questions