st: Re: twoway contour problem

[André Ferreira Coelho <andre.f.coelho2011@novasbe.pt>]

Dear statalist,

The error message i got (even after clearing missing values) is:


. twoway contour iratio country product
        _rbf_tps_setup():  3900  unable to allocate real
<tmp>[88118,88118]
          _rbf_tps_wrk():     -  function returned error
              _rbf_tps():     -  function returned error
                  _rbf():     -  function returned error
        _contour_setup():     -  function returned error
       _contour_filled():     -  function returned error
                 <istmt>:     -  function returned error


[M-2]   error . . . . . . . . . . . . . . . . . . . . . . . . Return code
3900
        out of memory
        Mata is out of memory; the operating system refused to supply
        what Mata requested.  There is no Mata or Stata setting that
        affects this, and so nothing in Mata or Stata to reset in order
        to get more memory.  You must take up the problem with your
        operating system.



Not sure if matsize solves anything given that for the version I am using
(Stata 12 SE) matsize upper limit is 11,000.
I am not sure if I'm not missing anything because what I would really
wanted in the end is a "matrix" of 1070*164 which is not even close to the
stata matsize limit.

Thank you,

Andre

-------- Original Message --------
Subject: twoway contour problem
Date: Mon, 14 Jan 2013 23:08:17 +0100
From: André Ferreira Coelho <andre.f.coelho2011@novasbe.pt>
To: <statalist@hsphsun2.harvard.edu>

Dear all, 

I am trying to generate a twoway countour graph using -twoway contour var1
var2 var3-. 

My data is in long form and was originally a panel dataset that I
restricted in order to contain only observations for a given year. It
happens that I my panel has 1070 products and 164 countries which sums up
to 175480 observations that would generate a 1070*164 matrix in the twoway
countour. 

Running the contour command it appears a 3900 error message stating lack
of memory. 

I tried then to eliminate missing values which would give me about 86000
observations. Once again the same problem of memory appeared. 

Does anyone know if it is possible to tackle this problem without
resorting to matlab or other software? 

Appreciated if you can help, 

André
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/