R help - Oct 2024 - Outputting a difftime object in PASTE or AS.CHARACTER is missing the unit + expand the format options?

Hello,

I was working on an R script using the datediff object, to log certain durations
in the data processing.
I ran in to the issue that outputing a datediff object using PASTE will leave
out the time unit by default.

When you paste() or as.character() a datediff object it only outputs the numeric
value, for example "1.23", so without the time unit secs, minutes,
hours etc. This is arguably is not very useful, because 1.23 seconds is very
different from 1.23 hours.

```
start_time <- Sys.time()
result <- sum(1:10000)
dur_time <- Sys.time() - start_time

print(paste(Sys.time(), "the process duration:", dur_time)) #
"2024-10-23 09:57:10.69578 the process duration: 0.00280904769897461"

print(dur_time)               # "Time difference of 0.002809048 secs"
print(paste(dur_time))        # "0.00280904769897461"
print(as.character(dur_time)) # "0.00280904769897461"
print(format(dur_time))       # "0.002809048 secs"

print(paste(dur_time, attr(dur_time, "units"))) #
"0.00280904769897461 secs"
```

So my questions are:
1)  is it possible to have the default output for as.character for the datediff
object also include the time unit? That way, it automatically outputs for
example "1.23 seconds" when used in combination with paste().
If someone really need just the numeric value without the unit, you can already
use as.numeric(dur_time).
2) Is it possible to expand the format() options for the datediff object, so
that it also accepts a strptime-like string, so like %H:%M:%S ? That way you
have more output options without having to rely on third-party libraries

regards,
Bas de reuver (UMCG, The Netherlands)
________________________________
 De inhoud van dit bericht is vertrouwelijk en alleen bestemd voor de
geadresseerde(n). Anderen dan de geadresseerde(n) mogen geen gebruik maken van
dit bericht, het niet openbaar maken of op enige wijze verspreiden of
vermenigvuldigen. Het UMCG kan niet aansprakelijk gesteld worden voor een
incomplete aankomst of vertraging van dit verzonden bericht.

The contents of this message are confidential and only intended for the eyes of
the addressee(s). Others than the addressee(s) are not allowed to use this
message, to make it public or to distribute or multiply this message in any way.
The UMCG cannot be held responsible for incomplete reception or delay of this
transferred message.

?s 09:11 de 23/10/2024, Reuver, B de (epi) via R-help
escreveu:
> Hello,
> 
> I was working on an R script using the datediff object, to log certain
durations in the data processing.
> I ran in to the issue that outputing a datediff object using PASTE will
leave out the time unit by default.
> 
> When you paste() or as.character() a datediff object it only outputs the
numeric value, for example "1.23", so without the time unit secs,
minutes, hours etc. This is arguably is not very useful, because 1.23 seconds is
very different from 1.23 hours.
> 
> ```
> start_time <- Sys.time()
> result <- sum(1:10000)
> dur_time <- Sys.time() - start_time
> 
> print(paste(Sys.time(), "the process duration:", dur_time)) #
"2024-10-23 09:57:10.69578 the process duration: 0.00280904769897461"
> 
> print(dur_time)               # "Time difference of 0.002809048
secs"
> print(paste(dur_time))        # "0.00280904769897461"
> print(as.character(dur_time)) # "0.00280904769897461"
> print(format(dur_time))       # "0.002809048 secs"
> 
> print(paste(dur_time, attr(dur_time, "units"))) #
"0.00280904769897461 secs"
> ```
> 
> So my questions are:
> 1)  is it possible to have the default output for as.character for the
datediff object also include the time unit? That way, it automatically outputs
for example "1.23 seconds" when used in combination with paste().
> If someone really need just the numeric value without the unit, you can
already use as.numeric(dur_time).
> 2) Is it possible to expand the format() options for the datediff object,
so that it also accepts a strptime-like string, so like %H:%M:%S ? That way you
have more output options without having to rely on third-party libraries
> 
> regards,
> Bas de reuver (UMCG, The Netherlands)
> ________________________________
>   De inhoud van dit bericht is vertrouwelijk en alleen bestemd voor de
geadresseerde(n). Anderen dan de geadresseerde(n) mogen geen gebruik maken van
dit bericht, het niet openbaar maken of op enige wijze verspreiden of
vermenigvuldigen. Het UMCG kan niet aansprakelijk gesteld worden voor een
incomplete aankomst of vertraging van dit verzonden bericht.
> 
> The contents of this message are confidential and only intended for the
eyes of the addressee(s). Others than the addressee(s) are not allowed to use
this message, to make it public or to distribute or multiply this message in any
way. The UMCG cannot be held responsible for incomplete reception or delay of
this transferred message.
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
https://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
Hello,

There are, in base R, methods

as.list.difftime
as.double.difftime

but not as.character.difftime.
If you write your own, it will be a functionality that only exists in 
your system and this can sometimes be a source for unexpected behavior.



as.character.difftime <- function(x) {
   u <- units(x)
   paste(unclass(x), u)
}

start_time <- Sys.time()
result <- sum(1:10000)
dur_time <- Sys.time() - start_time

as.character(dur_time)
#> [1] "0.000754117965698242 secs"


Hope this helps,

Rui Barradas


-- 
Este e-mail foi analisado pelo software antiv?rus AVG para verificar a presen?a
de v?rus.
www.avg.com

>>>>> Reuver, B de \(epi\) via R-help 
>>>>>     on Wed, 23 Oct 2024 08:11:50 +0000 writes:
    > Hello,
    > I was working on an R script using the datediff object, to log certain
durations in the data processing.
    > I ran in to the issue that outputing a datediff object using PASTE will
leave out the time unit by default.

    > When you paste() or as.character() a datediff object it only outputs
the numeric value, for example "1.23", so without the time unit secs,
minutes, hours etc. This is arguably is not very useful, because 1.23 seconds is
very different from 1.23 hours.

    > ```
    > start_time <- Sys.time()
    > result <- sum(1:10000)
    > dur_time <- Sys.time() - start_time

Note that  print(  ... )   below is unneeded:
 in R's toplevel, "auto" print() is (almost) always called

    > print(paste(Sys.time(), "the process duration:", dur_time)) #
"2024-10-23 09:57:10.69578 the process duration: 0.00280904769897461"

    > print(dur_time)               # "Time difference of 0.002809048
secs"
    > print(paste(dur_time))        # "0.00280904769897461"
    > print(as.character(dur_time)) # "0.00280904769897461"
    > print(format(dur_time))       # "0.002809048 secs"

i.e.,  the format()  method is doing the right thing, as it should.

The as.character() method shouldn't (and will not) be changed here.


    > print(paste(dur_time, attr(dur_time, "units"))) #
"0.00280904769897461 secs"
    > ```

    > So my questions are:

    > 1)  is it possible to have the default output for as.character for the
datediff object also include the time unit? That way, it automatically outputs
for example "1.23 seconds" when used in combination with paste().

No, see above.  You should use format() in such situations, not
just for "difftime" objects.

    > If someone really need just the numeric value without the unit, you can
already use as.numeric(dur_time).

indeed.
But as  as.character()  is defined (typically / in base R) to get rid of
attributes.

    > 2) Is it possible to expand the format() options for the datediff
object, so that it also accepts a strptime-like string, so like %H:%M:%S ? That
way you have more output options without having to rely on third-party libraries

That would be possible, but I'm not at all sure that it's
desirable / necessary  .. (but opinions may differ, I may change
mine after some good discussion...).

Did you look at the  ?difftime  help page  and its 'Examples' ?

I think everything you need is almost available.
The formatting in %H:%M .. etc is not directly possible,
and I think for good reasons:
Somewhat experienced R users can currently quickly see the
difference between  Date, Date-Times, and "difftime" objects
from their print output,  and I think that is desirable.


start_time <- structure(1729777, class = c("POSIXct",
"POSIXt"))
result <- sum(1:10000)
nowTime <- structure(1729888.123123123, class = c("POSIXct",
"POSIXt"))
##                          ^^^^^^^^^^ to see precision (not soo accurate with
POSIXct!)

(dur_time <- nowTime - start_time) # -> print
## Time difference of 1.852052 mins
format(dur_time)            # --> "1.852052 mins"       -- MM:
isn't that good!
format(dur_time, digits=12) # --> "1.85205205205 mins"

## You can switch from minutes to seconds :
dtime <- dur_time; units(dtime) <- "secs"
dtime ## Time difference of 111.1231 secs
format(dtime, digits = 10)
## [1] "111.1231231 secs"