Re: PDF calculation for a grid node

Can you provide me with the data file? It's hard to debug your script without this.

http://www.ncl.ucar.edu/report_bug.shtml#HowToFTP

I recommend that you upgrade your version of NCL, although this likely won't help you with your current issues.

--Mary

On Jul 15, 2013, at 7:19 AM, Ioannis Koletsis wrote:

> Dear Mary,
> first of all, thanks for your reply….
> Unfortunately, I cannot use function reshape because of my ncl version (6.0.0-beta) as the reshape can be used only in version 6.1.0 and later…
>
> However, I try to reshape my variable using statement new, according to the attached script….
> But, again the calculation of pdf for each grid point of my area has been failed….
>
> The following message was appeared:
> Number of elements of dimension (0) of argument (1) is (323) in function (pdfx), expected (1) elements….
>
> It seems that pdfx function works only for 1-D variable, but however I wonder if there is any way to calculate the pdf for each grid point in a selected area?
>
> Have you got any other idea?
>
> Thanks in advance
>
> Ioannis
>
> Attached Script
>
> ;*************************************************
> ; Calculation of PDF
> ;*************************************************
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
>
> begin
>
> DataDir = "/data3/koletsis/Current/A2/"
> Files = systemfunc ( " ls " + DataDir + \
> "C4IRCA3_A2_ECHAM5_DM_*.nc")
> numFiles = dimsizes(Files)
> print(Files)
> print(" ")
>
> f1 = addfiles(Files,"r")
>
> time = f1[:]->time
> ws = f1[:]->wss
> ws@_FillValue = 1.e+30
>
> lat2d = f1[0]->lat
> lon2d = f1[0]->lon
>
> time2 = ut_calendar(time, 0)
> time3 = ut_calendar(time, -2) ; Date format
>
> year = tointeger(time2(:,0)) ; Convert to integer
> month = tointeger(time2(:,1))
> day = tointeger(time2(:,2))
>
> beg_year = 1961
> end_year = 1990
> beg_month = 01
> end_month = 12
> beg_day = 01
> end_day = 31
>
> ymdStrt = beg_year*10000+beg_month*100+beg_day
> ymdLast = end_year*10000+end_month*100+end_day
>
> print("ymdStrt="+ymdStrt+" ymdLast="+ymdLast)
>
> itime = ind(time3.ge.ymdStrt.and.time3.le.ymdLast)
>
> if (ismissing(itime(0))) then ; Error message
> print("itime is missing:something is going wrong")
> exit
> end if
>
> ;======================================================================
> ; Select the subregion (Mediterranean Area)
> ;======================================================================
>
> ; latS = 30
> ; latN = 46
> ; lonW = -10
> ; lonE = 41
> latS = 36
> latN = 40
> lonW = 22
> lonE = 26
>
> ji = region_ind(lat2d,lon2d,latS,latN,lonW,lonE)
>
> jStrt = ji(0)
> jLast = ji(1)
> iStrt = ji(2)
> iLast = ji(3)
>
> llat2d = lat2d(jStrt:jLast,iStrt:iLast)
> llon2d = lon2d(jStrt:jLast,iStrt:iLast)
>
> nlat = dimsizes(llat2d(:,0))
> nlon = dimsizes(llon2d(0,:))
> nlev = dimsizes(ws(0,:,0,0))
>
> ws3 = ws(itime,:,jStrt:jLast,iStrt:iLast)
> ws2 = new((/nlev,nlat*nlon/),typeof(ws3))
> ws1 = 1.2697*ws2 ; wind speed at hub height
>
> print(ws1)
> ; printVarSummary(ws2)
> ; printVarSummary(ws1)
>
> ;======================================================================
> ; Calculate probability function
> ;======================================================================
>
> opt = True
> opt@bin_spacing = 0.2
>
> max_ws1 = dim_max_n(ws1,0)
> min_ws1 = 0.
>
> nbins = floattointeger(((max_ws1 - min_ws1) / 0.2)+1) ; bins number
>
> wspdf=(pdfx(ws1,nbins,opt))
>
>
> print(wspdf)
>
> end
>
>
>
>
> From: ncl-talk-bounces@ucar.edu [mailto:ncl-talk-bounces@ucar.edu] On Behalf Of Mary Haley
> Sent: Saturday, July 13, 2013 1:35 AM
> To: Ioannis Koletsis
> Cc: ncl-talk@ucar.edu
> Subject: Re: PDF calculation for a grid node
>
> I'm not sure what you mean by "grid node" versus "grid point".
>
> Did you try callilng the function without any loops?
>
> wspdf=(pdfx(ws1,nbins,opt))
>
> You may want "ws2" to be a 2D array rather than a 3D array, if you are trying to get the prob density across a lat/lon area for each level:
>
> llat2d = lat2d(jStrt:jLast,iStrt:iLast)
> llon2d = lon2d(jStrt:jLast,iStrt:iLast)
> nlat = dimsizes(lat2d(:,0))
> nlon = dimsizes(lat2d(0,:))
> nlev = dimsizes(ws(0,:,0,0))
> ws2 = reshape(ws(itime,:,jStrt:jLast,iStrt:iLast),(/nlev,nlat*nlon/))
> ws1 = 1.2697*ws2
>
> Then I think you can call pdfx without having to loop.
>
> --Mary
>
>
> On Jul 11, 2013, at 8:11 AM, Ioannis Koletsis <koletsis@noa.gr> wrote:
>
>
> Dear ncl users,
>
> I would like to calculate the pdf of wind speed for a grid node…
> When I calculated it for a grid point everything went well.
> The number of bins, and the maximum wind speed for each grid point have been estimated correctly (line number 92, 89 respectively), for each grid point throughout the time period (10957 records)….
> Moreover, the printVarSummary for max_ws1 and nbins give me the dimensions….
> Number of dimensions 3 (1 x 19 x 17) -> (height x rlat x rlon)
> As you can see at the attached script, I tried to solve the problem using do loops but unfortunately without success…
>
> Has anyone got any idea, how to calculate the pdf for each grid point….
> Any help would be appreciated….
>
> Ioannis
>
>
>
> ;*************************************************
> ; PDF Calculation
> ;*************************************************
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
>
> begin
>
> DataDir = "/data3/koletsis/Current/A2/"
> Files = systemfunc ( " ls " + DataDir + \
> "C4IRCA3_A2_ECHAM5_DM_*.nc")
> numFiles = dimsizes(Files)
> print(Files)
> print(" ")
>
> f1 = addfiles(Files,"r")
>
> time = f1[:]->time
> ws = f1[:]->wss
> ws@_FillValue = 1.e+30
>
> lat2d = f1[0]->lat
> lon2d = f1[0]->lon
>
> time2 = ut_calendar(time, 0)
> time3 = ut_calendar(time, -2) ; Date format
>
> year = tointeger(time2(:,0)) ; Convert to integer
> month = tointeger(time2(:,1))
> day = tointeger(time2(:,2))
>
> beg_year = 1961
> end_year = 1990
> beg_month = 01
> end_month = 12
> beg_day = 01
> end_day = 31
>
> ymdStrt = beg_year*10000+beg_month*100+beg_day
> ymdLast = end_year*10000+end_month*100+end_day
>
> print("ymdStrt="+ymdStrt+" ymdLast="+ymdLast)
>
> itime = ind(time3.ge.ymdStrt.and.time3.le.ymdLast)
>
> if (ismissing(itime(0))) then ; Error message
> print("itime is missing:something is going wrong")
> exit
> end if
>
> ;======================================================================
> ; Select the subregion (Mediterranean Area)
> ;======================================================================
>
> ; latS = 30
> ; latN = 46
> ; lonW = -10
> ; lonE = 41
> latS = 36
> latN = 40
> lonW = 22
> lonE = 26
>
> ji = region_ind(lat2d,lon2d,latS,latN,lonW,lonE)
>
> jStrt = ji(0)
> jLast = ji(1)
> iStrt = ji(2)
> iLast = ji(3)
>
> llat2d = lat2d(jStrt:jLast,iStrt:iLast)
> llon2d = lon2d(jStrt:jLast,iStrt:iLast)
>
> ws2 = ws(itime,:,jStrt:jLast,iStrt:iLast)
> ws1 = 1.2697*ws2 ; wind speed at hub height
>
> ; print(ws1)
> ; printVarSummary(ws2)
> ; printVarSummary(ws1)
>
> ;======================================================================
> ; Calculate probability function
> ;======================================================================
>
> opt = True
> opt@bin_spacing = 0.2
>
> max_ws1 = dim_max_n(ws1,0)
> min_ws1 = 0.
>
> nbins = floattointeger(((max_ws1 - min_ws1) / 0.2)+1) ; bins number
>
> ; print(nbins)
>
> dimx = dimsizes(ws1)
> nlev = dimx(0)
> nlat = dimx(1)
> mlon = dimx(2)
>
> do kk = 1, 10957
> do nl = 0, nlat-1
> do ml = 0, mlon-1
> do nt = 0, nlev-1
> wspdf=(pdfx(ws1(kk,nt,nl,ml),nbins,opt))
> end do
> end do
> end do
> end do
> print(wspdf)
>
> end
>
>
>
> <pdf.ncl>_______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
> No virus found in this message.
> Checked by AVG - www.avg.com
> Version: 2013.0.2904 / Virus Database: 3204/6490 - Release Date: 07/14/13
> <mary.ncl>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Jul 15 10:54:58 2013