[ncl-talk] Reopen: "proleptic_gregorian" Support in Date Routines

Karl-Hermann Wieners karl-hermann.wieners at mpimet.mpg.de
Tue Dec 1 07:43:08 MST 2015


Hi all,

as of NCL 6.3.0, cd_calendar cannot not handle time axes with the
calendar="proleptic_gregorian" attribute, despite what is written in the
docs. I've unearthed Albert and Mary's thread from January 2014 (see
below), where you planned to fix this issue in 6.2. Can you tell me what
happened to that or if and when this will be fixed?

Cheers,
Karl

--------------------------------------------------------------------

Re: "proleptic_gregorian" Support in Date Routines

    This message: [ Message body ] [ More options ]
    Related messages: [ Next message ] [ Previous message ] [ In reply to ]

From: Mary Haley <haley_at_nyahnyahspammersnyahnyah>
Date: Thu Jan 30 2014 - 10:48:45 MST

Sorry Albert, I missed the fact that you had attached a simple fix! I'll
look at this now and see if we can incorporate it into 6.2.0.

--Mary

On Jan 30, 2014, at 10:44 AM, Mary Haley <haley at ucar.edu> wrote:

> Hi Albert,
>
> Can you tell us what the different units are that you're seeing in
this one model, and how you are having to edit the files and resave them?
>
> I'm not exactly sure which "rebasing" feature you are referring to in
V6.2.0.
>
> Would the "calendar_decode2" function get you the support you need for
proleptic_gregorian?
>
>
http://www.ncl.ucar.edu/Document/Functions/User_contributed/calendar_decode2.shtml
>
>
> --Mary
>
> On Jan 28, 2014, at 8:38 AM, Albert Yau <albert.yau at stonybrook.edu> wrote:
>
>> Hi everyone,
>>
>> I found that the date routines in 6.1.2 are not yet fully supporting
the "proleptic_gregorian" calendar. For example, days_in_month will
return "illegal calendar" for this input. I have this problem while
processing some of the CMIP5 model's data. It is a trivial fix in
datesW.c which I have attached here, I hope it has already been included
in the next release.
>>
>> I have to say thanks for all of you developers, I am really looking
forward to the automatic time rebasing feature in 6.2.0, because one
particular model has monthly netcdf data with different units for a
hundred years. Right now I have to edit and save all the nc files before
calling addfiles.
>>
>> Best regards,
>> Albert
>> <ncl-6.1.2-calendar.patch>_______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk


More information about the ncl-talk mailing list