Re: ttest and gsn_csm_pres_hgt

From: Dennis Shea <shea_at_nyahnyahspammersnyahnyah>
Date: Tue Jun 18 2013 - 13:07:10 MDT

Since the files you are using are too large for our ftp,
The attached has not been debugged.

It includes a lot of 'printMinMax'.
For big arrays these can slow the code.
However, for debug purposes, they can halp
make sure the numbers look reasonable.
I would comment these out after you are confident
in the results.

         ;aps_tmp = in->aps_ave ; TIME, LAT, LON
          aps_tmp = in->aps_ave(0:35,:,:) ; FAST DEBUG

Also, for fast debug, I suggest you read in only
  small subset of the data (here 3 years worth)
After you are confident, comment the dbug and activate
the fukk array code.

Good luck

On 06/18/2013 08:46 AM, Dennis Shea wrote:
> @Ben: You should write to wrfhelp@ucar.edu
> They may be able to offer suggestions.
> ============================================
>
> Lena ...
>
> There are aspects of the NCL script that confuse me.
>
> [1] When you import a variable and then invert the grid via
> the ::-1 syntax, NCL is 'smart' about handling the
> associated coordinate meta data. Also, all meta data
> is 'private' to that variable. It is part of the variable's
> data structure. Look at the 'before' and 'after' ....
> Here the ::-1 reorders BOTH the array and the coordinate varialr
> to reflect the array reordering.
>
> aps_tmp = in->aps_ave ; TIME, LAT, LON
> printVarSummary(aps_tmp) ; note the 'lat' coordinate variable
> aps = aps_tmp(:,::-1,:)
> printVarSummary(aps_tmp) ; NCL has reordered the 'lat'
>
> U_wind_tmp = in->n->um1_avee ; TIME, LEV, LAT, LON
> printVarSummary(U_wind_tmp)
> U_wind = U_wind_tmp(:,:,::-1,:)
> printVarSummary(U_wind_tmp)
>
> Incidently, you can do the reorder without using a 'tmp' variable
>
> aps = in->aps_ave(:,::-1,:)
> U_wind = in->n->um1_ave(:,:,::-1,:)
>
>
> [2] Generally there is no reason to explicitly read coordinate variable
> of a file.
>
> time_EMAC = in->time
> lat_EMAC = in->lat
> lon_EMAC = in->lon
> lev_EMAC = in->lev
>
> When you do
> U_wind_tmp = in->um1_ave
> and
> aps_tmp = in->aps_ave
>
> each variable has its' own 'time'. Since they are from
> the same file, they are the same. Look at 'printVarSummary'
>
> [3] When you do
>
> aps&lat = aps&lat(::-1)
> and
> U_wind&lat = U_wind&lat(::-1)
>
> You have reordered the 'lat' coordinate variable BUT not the array
> with which it is associated. I would say this is incorrect.
>
> [4] u_ave = dim_avg_n_Wrap(newgrid(:,:,::-1,:),0)
>
> I assume you are reordering to the N->S on purpose
>
> I am at a mtg and must go ...
>
> You can ftp the source file to
>
> ftp ftp.cgd.ucar.edu
> anonymous
> <use your email address for the password>
> cd incoming
> put ...
>
> quit
>
> Send email to me after you have successfully completed the transfer.
>
> Regards
>
>
> On 6/18/13 7:02 AM, Lena Frey wrote:
>> Hi,
>> @Dennis: Please find attached my ncl script. Thank you.
>> @Ben: Sorry, I don't know anyone who has used ECHAM5 to drive WRF.
>> Lena
>>
>>
>>
>> _______________________________________________
>> 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

Received on Tue Jun 18 13:07:21 2013

This archive was generated by hypermail 2.1.8 : Mon Jun 24 2013 - 11:46:47 MDT