R help - Oct 2007 - error message: "the instruction at "0x6c72c676" referenced memory of "0x00000010". The memory could not be "written"."

Hello everyone,

I have a problem with a selfwritten R-program. 

Short description about the program:
It gets values via SQL from a MS-Access-database, then forms them to
matrices or dataframes and makes some calculations with them. At the end it
fills the results in a result table in a MS-Access-database via SQL.


Short description about the problem:
If the size is small (rows <= 200 , columns = 16, the programm needs less
than 30min), everything works great.
But if the size is big (rows > 200, columns = 16; the program needs more
than 30min), I always get this error message: 

"the instruction at "0x6c72c676" referenced memory of
"0x00000010". The
memory could not be "written"." 

This happens at different points in the progress of the program.

If anyone has an idea about this problem, I would be very grateful, if he
could help me.

Best regards

Freundliche Gr??e
Oliver H?nig
____________________________________________
Commerzbank AG 
Corporates & Markets
Research - Quantitative Portfolio Strategie
DLZ - Geb?ude 2, H?ndlerhaus
Mainzer Landstrasse 153
D-60327 Frankfurt
Phone:   +49 69 136 45178
Hoenig.Oliver at commerzbank.com
http://www.commerzbank.de <http://www.commerzbank.de> 
____________________________________________
COMMERZBANK Aktiengesellschaft, Frankfurt am Main
Registergericht/Registered Office: Amtsgericht Frankfurt am Main (HRB 32000)
Vorsitzender des Aufsichtsrates/Chairman of the Supervisory Board: Martin
Kohlhaussen
Vorstand/Board of Managing Directors: Klaus-Peter M?ller
(Sprecher/Chairman), Martin Blessing, Wolfgang Hartmann,
Achim Kassow, Bernd Knobloch, Michael Reuther, Eric Strutz, Nicholas Teller

Hello Oliver,

Recalling my past, this looks to me like a message one
gets when there is a memory violation in C.
When the problem size is small this violation may
"visit" a memory address which it is allowed to and
there is no exception (but even then you can not fully
trust your result). When the memory usage increases it
"visits" an address which it is not allowed to and the
exceptions occurs. After all, a big portion of R is
written in C and Fortran.
I do not know how R is supposed to deal with native
code and whether it should have prevented the
occurrence of such an exception.

Regards,

Moshe.

--- "Hoenig, Oliver" <Hoenig.Oliver at commerzbank.com>
wrote:
> Hello everyone,
> 
> I have a problem with a selfwritten R-program. 
> 
> Short description about the program:
> It gets values via SQL from a MS-Access-database,
> then forms them to
> matrices or dataframes and makes some calculations
> with them. At the end it
> fills the results in a result table in a
> MS-Access-database via SQL.
> 
> 
> Short description about the problem:
> If the size is small (rows <= 200 , columns = 16,
> the programm needs less
> than 30min), everything works great.
> But if the size is big (rows > 200, columns = 16;
> the program needs more
> than 30min), I always get this error message: 
> 
> "the instruction at "0x6c72c676" referenced memory
> of "0x00000010". The
> memory could not be "written"." 
> 
> This happens at different points in the progress of
> the program.
> 
> If anyone has an idea about this problem, I would be
> very grateful, if he
> could help me.
> 
> Best regards
> 
> Freundliche Gr??e
> Oliver H?nig
> ____________________________________________
> Commerzbank AG 
> Corporates & Markets
> Research - Quantitative Portfolio Strategie
> DLZ - Geb?ude 2, H?ndlerhaus
> Mainzer Landstrasse 153
> D-60327 Frankfurt
> Phone:   +49 69 136 45178
> Hoenig.Oliver at commerzbank.com
> http://www.commerzbank.de
> <http://www.commerzbank.de> 
> ____________________________________________
> COMMERZBANK Aktiengesellschaft, Frankfurt am Main
> Registergericht/Registered Office: Amtsgericht
> Frankfurt am Main (HRB 32000)
> Vorsitzender des Aufsichtsrates/Chairman of the
> Supervisory Board: Martin
> Kohlhaussen
> Vorstand/Board of Managing Directors: Klaus-Peter
> M?ller
> (Sprecher/Chairman), Martin Blessing, Wolfgang
> Hartmann,
> Achim Kassow, Bernd Knobloch, Michael Reuther, Eric
> Strutz, Nicholas Teller
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained,
> reproducible code.
>