Re: Problems combining dim_num_n and dim_numrun_n to number of runs of a given length

Hi Dominique,
Continuing to think about what you are trying to do, I think you want to
use dim_sum_n instead of dim_num_n, as you wish to sum up the number of
times a specific criteria has been met.
http://www.ncl.ucar.edu/Document/Functions/Built-in/dim_sum_n.shtml
Adam

On 7/2/12 4:08 PM, Adam Phillips wrote:
> Hi Dominique,
> I think the issue has to do with how you are using dim_num_n, and that
> may be related to the fact that I t hink example #4 on the dim_num_n
> has an incorrect comment.
>
> Example #4 should say this:
> N =*dim_num_n*(z,(/2,3/)) ; N(ntime,klev)
> In other words, the dimensions that you specify in dim_num_n will be
> eliminated from the returned array. I will get the website fixed..
>
> You say this:
> tc2_tally0 = dim_num_n (uk_cold0, (/3,4/))
> That statement though will eliminate your latitude and longitude
> arrays. I think you might want to say:
> tc2_tally0 = dim_num_n (uk_cold0, (/0,1,2/))
> Note though that the above statement will eliminate your first three
> dimensions from uk_cold0, and I am not sure if you want to sum up
> across all of the first 3 dimensions of uk_cold0 and uk_cold1.
>
> I would advise that you check the output from dim_num_n to make sure
> it is doing what you think it is doing.
>
> If that answer does not help, please let ncl-talk know.. Adam
>
> I
>
> On 7/1/12 12:03 PM, Cannavina, Dominique wrote:
>>
>> Hi NCL talk
>>
>> I am trying to use dim_num_n to sum up the number of runs of data (in
>> this case temp below tcrit =0degC) which have been produced from
>> dim_numrun_n. My data extends over europe so I need to extract data
>> at all grid points. The data files have 3 models, each model with 9
>> ensemble runs, and each ensemble run with 180 days of min
>> temperature. I am just looking at 1 model at the moment.
>>
>> I am trying to use the dim_num_n to count the number of each
>> 'duration of run' which has been produced by dim_numrun_n
>>
>> When I try and use a do loop and dim_num_n the function cuts off the
>> lat and lon parts of the variable so I end up with [num file |
>> ensemble | step ] which then causes me problems when I try and plot
>> the data later on.
>>
>> An exerpt of the main section is given below-but the full script and
>> a sample of the variables is also attached.
>>
>> Any advice would be much appreciated.
>>
>> Kind regards
>>
>> Dominique
>>
>> ; What times and how many time steps are in the data set?
>>
>> ntimes = dimsizes(times) ; number of times in the file
>>
>> ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
>>
>> ;First look at data from model ukmo
>>
>> ; Get the variables we will need
>>
>> t2_uk = ukmo ; T2 in Kelvin
>>
>> tc2_uk = t2_uk-273.16 ; T2 in C
>>
>> ;Counts the number of consecutive days that the temperature falls to
>> or below the stipulated temperature.
>>
>> ; specify cut off temperature 'tcrit'
>>
>> ;I am trying to tally the number of cold spells per set of data (i.e.
>> for each ensemble for each model for each year/start of ensemble run)
>> but I need to be able to count how many 1,2,3....35 day duration
>> events occur
>>
>> ;ukmo(ncl_join,step,number,latitude,longitude) => ukmo(0,1,2,3,4)
>>
>> tStrt = 0
>>
>> tLast = 179
>>
>> tSkip = 1
>>
>> ; temporary variable
>>
>> tc2 = tc2_uk(:,tStrt:tLast:tSkip,:,:,:) ; convenience/clarity
>>
>> printVarSummary(tc2) ; does this look right?
>>
>> ;tc2 [ncl_join | 2] x [step | 180] x [number | 9] x [latitude | 18] x
>> [longitude | 30]
>>
>> ;make sure dimensions of array fit with expected dimensions given
>> above by specifying the following
>>
>> dimt2 = dimsizes(tc2)
>>
>> nfils = dimt2(0)
>>
>> ntim = dimt2(1)
>>
>> nens = dimt2(2)
>>
>> nlat = dimt2(3)
>>
>> mlon = dimt2(4)
>>
>> tcrit = 0
>>
>> uk_cold = where(tc2.le.tcrit, 1, 0) ; 0/1 array
>>
>> printVarSummary(uk_cold)
>>
>> uk_cold0 = dim_numrun_n(uk_cold, 0, 0) ; opt = 0 means total
>>
>> uk_cold1 = dim_numrun_n(uk_cold, 1, 0) ; opt = 1 means unique
>>
>> uk_cold0@long_name = "total cold runs"
>>
>> uk_cold0@cold_crit_value = tcrit
>>
>> copy_VarCoords(tc2, uk_cold0)
>>
>> uk_cold1@long_name = "unique cold runs"
>>
>> uk_cold1@cold_crit_value = tcrit
>>
>> copy_VarCoords(tc2, uk_cold1)
>>
>> ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
>>
>> ;Second and third models to be added later
>>
>> ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
>>
>> ;I now want to produce plots showning the average number of times in
>> a given period there has been a 1 day duration cold event.... max day
>> duration cold event
>>
>> ;so if the max duration of the cold event is 20 days, I will have 20
>> plots. I have 3 models each with 9 ensembles so for this period
>> tally/(3*9)
>>
>> ; Not been able to produce this yet so try and get following two outputs
>>
>> pltres = True
>>
>> ;pltres@PanelPlot = True
>>
>> plots = new (60, graphic)
>>
>> nuk_cold = dimsizes(uk_cold)
>>
>> do n=0,nuk_cold(0)-1
>>
>> tc2_tally0 = dim_num_n (uk_cold0, (/3,4/))
>>
>> tc2_tally1 = dim_num_n (uk_cold1, (/3,4/))
>>
>> -
>>
>> -
>>
>> -script defining boundaries lat/lon etc
>>
>> -
>>
>> -
>>
>> plots(n) = gsn_csm_contour_map(wks,tc2_tally0(n,0,0,:,:),res0)
>>
>> plots(n) = gsn_csm_contour_map(wks,tc2_tally1(n,0,0,:,:),res1)
>>
>> end do
>>
>> end
>>
>>
>>
>> _______________________________________________
>> 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 Mon Jul 2 16:26:13 2012