[ncl-talk] Is function trend_manken_n available?

[Dennis Shea]

I was about to delete this from the queue but thought I'd add something.

FYI only:

If the residuals of the line derived via simple linear regression (SLR) are
normally distributed,
there is virtually no difference between the regression and the M-K
(Thiel-Sen) estimates.

See attached which shows the SLR and M-K (Thiel-Sen) fits for the 1st half;
2nd half and whole series.

Again: the NCL Mann-Kendall (Thiel-Sen) function will be available in 6.3.0

time series of global anomalies:
http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/

The data are annual means and reside in column 2 of the input file. The
file structure is described at:
http://hadobs.metoffice.com/hadcrut3/diagnostics/time-series.html



On Fri, Sep 5, 2014 at 2:41 PM, Xi Chang <xi.chang01 at gmail.com> wrote:

> sorry i missed the fortran code.
> i found  the significant test in the original code.
>
> Thanks
>
>
>
> On Friday, September 5, 2014, Xi Chang <xi.chang01 at gmail.com> wrote:
>
>> and how to test whether the trend is significant at 95% or not?
>>
>> Thanks
>>
>>
>>
>> On Friday, September 5, 2014, Ruping Mo <rupingmo at gmail.com> wrote:
>>
>>> Thank you very much, Mary. I will give this a try. I also want to thank
>>> Dennis for his effort.
>>>
>>> Ruping
>>>
>>> On Friday, September 5, 2014, Mary Haley <haley at ucar.edu> wrote:
>>>
>>>> Ruping,
>>>>
>>>> On occasion, we will create documentation for upcoming functions, even
>>>> though they are not available yet.  We always try to put the version number
>>>> on new functions, so you know if your version of NCL has that function or
>>>> not.
>>>>
>>>> We have not actually implemented the wrapper for this function yet, but
>>>> Dennis had provided the Fortran code, a sample NCL script, and a sample
>>>> dataset.
>>>>
>>>> If you want to give this a try, then download the three attached files,
>>>> then type:
>>>>
>>>> WRAPIT kens_trend_dp.ncl.f
>>>> ncl test.ken_trend.ncl
>>>>
>>>> You should see output like this:
>>>>
>>>> (0) ================ TA ==================================
>>>>  n=165 s= 8059 nc= 13530  nt=165 var=503597.0 vartie=    19.7 trend=
>>>> 0.00436
>>>> (0) stau=8059
>>>> (0)    z=11.35496228080661
>>>> (0) prob =1
>>>> (0) trend=0.004361904761904762
>>>>
>>>> If you have questions about this, please post back to ncl-talk.
>>>>
>>>> --Mary
>>>>
>>>>
>>>>
>>>> On Thu, Sep 4, 2014 at 2:06 PM, Ruping Mo <rupingmo at gmail.com> wrote:
>>>>
>>>>> Dear NCL,
>>>>>
>>>>> I found this new function in the function list. Is it available now,
>>>>> or in version 6.3 only?
>>>>>
>>>>> Regards,
>>>>>
>>>>> Ruping
>>>>>
>>>>> _______________________________________________
>>>>> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140908/b595d04f/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: trend_manken.png
Type: image/png
Size: 46274 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20140908/b595d04f/attachment.png