[ncl-talk] Sequences of values

[Rashed Mahmood]

How does your script look like? and what is the size of input data?

On Mon, Mar 26, 2018 at 8:45 AM, Kiswendsida Hyacinthe GUIGMA <
karongseba at gmail.com> wrote:

> Hi Rashed and everyone,
> Thank you for your suggestion.
> I also had the idea of using many loops. It works perfectly for no large
> data. But when it comes to apply it to a 3D data covering half of the
> Africa domain, it slows down the execution and ends up killing the process.
> That's why I was thinking that there exist an other way to do that using
> not as much iterations.
>
> Best
>
>
> -----------------------------
> GUIGMA
>
>
> On Saturday, 24 March 2018, 18:30:38 GMT, Rashed Mahmood <
> rashidcomsis at gmail.com> wrote:
>
>
> Hi Guigma,
> For 1 dimensional series you can use the following. however, you would
> need to adapt this to your 3D data sets through loops/ind function and need
> to do TESTING, this is a very time consuming process.
> begin
> ;  X1 = (/0,1,0,0,1,1,1,1,1,0,0,1,0,1,1,1,1,0,0,1,0/) ; orignal
>    X1 = (/0,1,0,0,1,1,1,1,1,0,0,1,0,1,1,1,1,0,0,1,0,1,1,0,0,1,0,1,0,1,1/)
> ; for testing
>
>    X2 = new(dimsizes(X1),integer)
>
>  do n=0,dimsizes(X1)-1
>      if(n.eq.0)
>        if(X1(n).eq.1.and.X1(n+1).eq.0)
>          X2(n) = 0
>         else
>          X2(n) = X1(n)
>        end if
>      end if
>
>     if(n.gt.0.and.n.lt.dimsizes(X1)-1)
>       if(X1(n).eq.1)
>          if(X1(n-1).eq.0.and.X1(n+1).eq.0)
>             X2(n) = 0
>           else
>             X2(n) = X1(n)
>          end if
>        else
>           X2(n) = X1(n)
>       end if
>
>     end if
>
>     if(n.eq.dimsizes(X1)-1)
>       if(X1(n).eq.1.and.X1(n-1).eq.1)
>         X2(n) = X1(n)
>        else
>         X2(n) = 0
>       end if
>     end if
>  end do
>     print(X1+" ... "+X2)
>
> end
>
>
> Note I have NOT tested it as much one should so I would let you to test
> and find out if that is what you are looking for.
>
> Cheers,
> Rashed
>
>
>
>
>
>
>
> On Sat, Mar 24, 2018 at 6:44 AM, Kiswendsida Hyacinthe GUIGMA <
> karongseba at gmail.com> wrote:
>
> Thank you for your reply and for your help.
> Actually, before to post my question, I tried to make a use of
> dim_numrun_n but it didn't solve my problem.
>
> You misunderstood a bit my [1]. I said that I successfully did that... So
> it is OK for that.
>
> For [2] , what the dim_numrun_n  function allows is just to do a count.
> What me I'd rather like is to set to _FillValue (or to 0 if use of where
> function) all values that are not in a sequence of at least 04 days. As an
> example I would like to have something as follows:
> X1=(/0,1,0,0,1,1,1,1,1,0,0,1, 0,1,1,1,1,0,0,1,0,/) ==========> X2= (/0,*0*
> ,0,0,1,1,1,1,1,0,0,*0*,0,1, 1,1,1,0,0,*0*,0,/).
> In other words it is to replace all "isolated" 1s by 0.
>
> Kindest regards
>
>
>
>
>
> -----------------------------
> GUIGMA
>
>
> On Friday, 23 March 2018, 22:17:11 GMT, Dennis Shea <shea at ucar.edu>
> wrote:
>
>
> Sorry, I'm I am not looking at any code.
>
>    tcrit = 30.
>
>    f  = addfile(...)
>   T = f->T                  ; (time,lat,lon)
>
>
> [1] "setting values below the threshold to _FillValue" is simple.
>
> https://www.ncl.ucar.edu/ Document/Functions/Built-in/ where.shtml
> <https://www.ncl.ucar.edu/Document/Functions/Built-in/where.shtml>
>
>    Tcrit = where(T.ge.tcrit, T, T at _FillValue)
>    copy_VarMeta(T, Tcrit)
> ===========
> [2] You do not need to do [1]
>
> Maybe you could use the following function:
>
> =========
> http://www.ncl.ucar.edu/ Document/Functions/Built-in/ dim_numrun_n.shtml
> <http://www.ncl.ucar.edu/Document/Functions/Built-in/dim_numrun_n.shtml>
> ****Look at the examples****
> =========
>
>   T10= where(T.ge.tcrit,1, 0)
>   T10 at long_name = "Exceed=1; No=0"
>   copy_VarCoords(T, T10)
>   printVarSummary(T10)
>
>  r0 = *dim_numrun_n* (T10, *0*, 0)  ; (ntim,nlat,mlon)
>  r1 = *dim_numrun_n* (T10, *1*, 0)
>  copy_VarCoords(T, r0)
>  copy_VarCoords(T, r1)
>
> ;---Pick one grid point to see what is happening
>
>   LAT = ..
>   LON = ..
>   print(T(:,{LAT},{LON})+"  "+Tcrit(:,{LAT},{LON})+"  "+T10(:,{LAT},{LON}) \
>                         +"  "r0(:,{LAT},{LON})+"   "+r1(:,{LAT},{LON})
>
>
>
> On Fri, Mar 23, 2018 at 3:29 PM, Kiswendsida Hyacinthe GUIGMA <
> karongseba at gmail.com> wrote:
>
> Dear NCL users,
> I have 3D (time,lat,lon) array of temperature from which I would like to
> keep only sequences (consecutive days) exceeding a certain threshold.  This
> consist concretely in setting values below the threshold to _FillValue.
> The length of sequence is set to 4. So far,  have been able  to set to
> _FillValu only individual days that exceed the threshold. Then I would be
> pleased if you could provide me with a method to also set to _FillValue
> values that are not into sequences of at least 04 days.
>
> Please find attached my data and my code.
>
> Regards
>
>
> -----------------------------
> GUIGMA
>
> ______________________________ _________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/ mailman/listinfo/ncl-talk
> <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
>
>
>
> ______________________________ _________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/ mailman/listinfo/ncl-talk
> <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20180326/a43eecf1/attachment.html>