[ncl-talk] how to allocate output variables for stat_medrng?

[Rick Brownrigg]

Hi Jian,

I believe you may have found a bug in the case for this function when the
output values should be a scalar.  I need to confirm with my colleagues,
and I'll file a ticket if so.  In the mean time, a work around is to make
the a,b,c,d variables the same dimension, and only use the 1st element of
each on return from the function:

x = fspan(1,10,100)
a = x
b = x
c = x
d = floattoint(x)
d(0) = 1
;; Note that a,b,c,d are the same dimension, and effectively contain
garbage values
stat_medrng(x,a,b,c,d)
print("median = " + a(0))
print("mid-value = " + b(0))
print("range = " + d(0))


Hope that helps...
Rick

On Thu, Apr 6, 2017 at 8:40 AM, Jian Zheng <jzheng8606 at gmail.com> wrote:

> Hi all,
>
> When I use procedure stat_medrng, I get some errors. Below are some simple
> testing code
>
> ncl 0> x=fspan(1,10,100)
> ncl 1> a=0.
> ncl 2> b=0.
> ncl 3> c=0.
> ncl 4> d=0
> ncl 5> stat2(x,a,b,d)
> ncl 6> stat_medrng(x,a,b,c,d)
> fatal:stat_medrng: The dimensions of xmedian, xrange, xmrange, and nptused
> must be the same as the left-most dimensions of x
> fatal:["Execute.c":8565]:Execute: Error occurred at or near line 6
> ncl 8> stat4(x,a,b,c,c,d)
>
> You can see it works well with stat2 or stat4, but not stat_medrng. How to
> define the output variables with stat_medrng for one-dimensional input?
>
> Thanks,
> Jian
>
> _______________________________________________
> ncl-talk mailing list
> ncl-talk at ucar.edu
> 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/20170406/19a766b9/attachment.html