R help - Apr 2017 - as.POSIXct character string is not in a standard unambiguous format

Hi Ben

Thanks for your answer
I have already tried this, as well as
     x <- as.POSIXct(strptime("2002-02-02 02:02", "%Y-%m-%d
%H:%M"))
It works! But it does not fix it widely for all tests used during the 
"make check" step at compile time. Unless I patch all of them.

There is something with localtime but I cannot find what.

On another machine with another Linux OS, and the same environmental 
variables
    x <- as.POSIXct("2002-02-02 02:02")
works fine.

S?bastien
> Hi,
>
> I can't answer the question about R 3.3.3, but I don't see anything
in the update notes.
>
> http://mirror.its.dal.ca/cran/doc/manuals/r-release/NEWS.html
>
> In the meantime, would it skirt your issue if you explicitly stated the
format?
>
> x <- as.POSIXct("2002-02-02 02:02", format = "%Y-%m-%d
%H:%M")
>
> Ben
>
>
>
>> On Apr 5, 2017, at 11:21 AM, Sebastien Moretti <sebastien.moretti at
unil.ch> wrote:
>>
>> Hi
>>
>> I have lots of issues when I try to install R 3.3.3 during the
"make check" step.
>>
>> Every time a call to as.POSIXct is done in test scripts, I got the same
error message:
>> e.g. x <- as.POSIXct("2002-02-02 02:02")
>> Error in as.POSIXlt.character(x, tz, ...) :
>>  character string is not in a standard unambiguous format
>>
>>
>> It looks to be linked to localtime but when I compiled R 3.3.2  6
months ago, the same test scripts were there and succeeded.
>>
>>
>> Is there an environmental variable to use to change the as.POSIXct
behavior?
>>
>> Regards
>>
>> --
>> S?bastien
>>
>> ______________________________________________
>> 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
http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
> Ben Tupper
> Bigelow Laboratory for Ocean Sciences
> 60 Bigelow Drive, P.O. Box 380
> East Boothbay, Maine 04544
> http://www.bigelow.org

I have just found the solution.

We have a custom Linux distribution that allows us to have several R (+ 
glibc and others) versions in parallel for tools related to our job domain.

We have another etc/ folder for those tools and R looks for the 
localtime file there, not in /etc/.
So linking /etc/localtime to our etc/ folder makes R happy and
     x <- as.POSIXct("2002-02-02 02:02")
works fine!

S?bastien
> Hi Ben
>
> Thanks for your answer
> I have already tried this, as well as
>     x <- as.POSIXct(strptime("2002-02-02 02:02",
"%Y-%m-%d %H:%M"))
> It works! But it does not fix it widely for all tests used during the
> "make check" step at compile time. Unless I patch all of them.
>
> There is something with localtime but I cannot find what.
>
> On another machine with another Linux OS, and the same environmental
> variables
>    x <- as.POSIXct("2002-02-02 02:02")
> works fine.
>
> S?bastien
>
>> Hi,
>>
>> I can't answer the question about R 3.3.3, but I don't see
anything in
>> the update notes.
>>
>> http://mirror.its.dal.ca/cran/doc/manuals/r-release/NEWS.html
>>
>> In the meantime, would it skirt your issue if you explicitly stated
>> the format?
>>
>> x <- as.POSIXct("2002-02-02 02:02", format =
"%Y-%m-%d %H:%M")
>>
>> Ben
>>
>>
>>
>>> On Apr 5, 2017, at 11:21 AM, Sebastien Moretti
>>> <sebastien.moretti at unil.ch> wrote:
>>>
>>> Hi
>>>
>>> I have lots of issues when I try to install R 3.3.3 during the
"make
>>> check" step.
>>>
>>> Every time a call to as.POSIXct is done in test scripts, I got the
>>> same error message:
>>> e.g. x <- as.POSIXct("2002-02-02 02:02")
>>> Error in as.POSIXlt.character(x, tz, ...) :
>>>  character string is not in a standard unambiguous format
>>>
>>>
>>> It looks to be linked to localtime but when I compiled R 3.3.2  6
>>> months ago, the same test scripts were there and succeeded.
>>>
>>>
>>> Is there an environmental variable to use to change the as.POSIXct
>>> behavior?
>>>
>>> Regards
>>>
>>> --
>>> S?bastien

You always need to set your timezone somehow when converting to POSIXt.
Technically the method for doing this varies by OS, but on all environments I
have worked with you can set the default timezone with something like

Sys.setenv( TZ="Etc/GMT+5" )

In your example, some timezones supporting daylight savings would regard that
time as invalid (clock "jumps forward"). If the default
("system") timezone on a particular server does not match the data you
are working with, use the above command to make it easy to work with data from
the desired timezone temporarily. Read ?timezones.
-- 
Sent from my phone. Please excuse my brevity.

On April 5, 2017 11:55:34 PM PDT, Sebastien Moretti <sebastien.moretti at
unil.ch> wrote:
>Hi Ben
>
>Thanks for your answer
>I have already tried this, as well as
>     x <- as.POSIXct(strptime("2002-02-02 02:02",
"%Y-%m-%d %H:%M"))
>It works! But it does not fix it widely for all tests used during the 
>"make check" step at compile time. Unless I patch all of them.
>
>There is something with localtime but I cannot find what.
>
>On another machine with another Linux OS, and the same environmental 
>variables
>    x <- as.POSIXct("2002-02-02 02:02")
>works fine.
>
>S?bastien
>
>> Hi,
>>
>> I can't answer the question about R 3.3.3, but I don't see
anything
>in the update notes.
>>
>> http://mirror.its.dal.ca/cran/doc/manuals/r-release/NEWS.html
>>
>> In the meantime, would it skirt your issue if you explicitly stated
>the format?
>>
>> x <- as.POSIXct("2002-02-02 02:02", format =
"%Y-%m-%d %H:%M")
>>
>> Ben
>>
>>
>>
>>> On Apr 5, 2017, at 11:21 AM, Sebastien Moretti
><sebastien.moretti at unil.ch> wrote:
>>>
>>> Hi
>>>
>>> I have lots of issues when I try to install R 3.3.3 during the
"make
>check" step.
>>>
>>> Every time a call to as.POSIXct is done in test scripts, I got the
>same error message:
>>> e.g. x <- as.POSIXct("2002-02-02 02:02")
>>> Error in as.POSIXlt.character(x, tz, ...) :
>>>  character string is not in a standard unambiguous format
>>>
>>>
>>> It looks to be linked to localtime but when I compiled R 3.3.2  6
>months ago, the same test scripts were there and succeeded.
>>>
>>>
>>> Is there an environmental variable to use to change the as.POSIXct
>behavior?
>>>
>>> Regards
>>>
>>> --
>>> S?bastien
>>>
>>> ______________________________________________
>>> 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
>http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>
>> Ben Tupper
>> Bigelow Laboratory for Ocean Sciences
>> 60 Bigelow Drive, P.O. Box 380
>> East Boothbay, Maine 04544
>> http://www.bigelow.org
>
>______________________________________________
>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
>http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.

I cannot imagine a less desirable solution.  This is the opposite of portable
programming.
-- 
Sent from my phone. Please excuse my brevity.

On April 6, 2017 5:29:08 AM PDT, Sebastien Moretti <sebastien.moretti at
unil.ch> wrote:
>I have just found the solution.
>
>We have a custom Linux distribution that allows us to have several R (+
>
>glibc and others) versions in parallel for tools related to our job
>domain.
>
>We have another etc/ folder for those tools and R looks for the 
>localtime file there, not in /etc/.
>So linking /etc/localtime to our etc/ folder makes R happy and
>     x <- as.POSIXct("2002-02-02 02:02")
>works fine!
>
>S?bastien
>
>> Hi Ben
>>
>> Thanks for your answer
>> I have already tried this, as well as
>>     x <- as.POSIXct(strptime("2002-02-02 02:02",
"%Y-%m-%d %H:%M"))
>> It works! But it does not fix it widely for all tests used during the
>> "make check" step at compile time. Unless I patch all of
them.
>>
>> There is something with localtime but I cannot find what.
>>
>> On another machine with another Linux OS, and the same environmental
>> variables
>>    x <- as.POSIXct("2002-02-02 02:02")
>> works fine.
>>
>> S?bastien
>>
>>> Hi,
>>>
>>> I can't answer the question about R 3.3.3, but I don't see
anything
>in
>>> the update notes.
>>>
>>> http://mirror.its.dal.ca/cran/doc/manuals/r-release/NEWS.html
>>>
>>> In the meantime, would it skirt your issue if you explicitly stated
>>> the format?
>>>
>>> x <- as.POSIXct("2002-02-02 02:02", format =
"%Y-%m-%d %H:%M")
>>>
>>> Ben
>>>
>>>
>>>
>>>> On Apr 5, 2017, at 11:21 AM, Sebastien Moretti
>>>> <sebastien.moretti at unil.ch> wrote:
>>>>
>>>> Hi
>>>>
>>>> I have lots of issues when I try to install R 3.3.3 during the
>"make
>>>> check" step.
>>>>
>>>> Every time a call to as.POSIXct is done in test scripts, I got
the
>>>> same error message:
>>>> e.g. x <- as.POSIXct("2002-02-02 02:02")
>>>> Error in as.POSIXlt.character(x, tz, ...) :
>>>>  character string is not in a standard unambiguous format
>>>>
>>>>
>>>> It looks to be linked to localtime but when I compiled R 3.3.2 
6
>>>> months ago, the same test scripts were there and succeeded.
>>>>
>>>>
>>>> Is there an environmental variable to use to change the
as.POSIXct
>>>> behavior?
>>>>
>>>> Regards
>>>>
>>>> --
>>>> S?bastien
>
>______________________________________________
>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
>http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.