Current Messages

Netlib Discussion Forum: Netlib: LAPACK: Current Messages
By
Lin Zhou on Wednesday, October 7, 1998 - 11:00 am:

When I use IMDIF, I find it is very sensitive to the initial guess. Are there any ways to deal with it? Thanks!


By Slavomir Matas on Tuesday, October 27, 1998 - 04:14 am:

Hello,

Does anyone compiled the CLAPACK2.0 using C/C++ WATCOM compiler 10.6 ?

I have tried to compile CLAPACK with WATCOM C/C++10.6 compiler on
PC running windows 95, but I have got a lots of messages
e.g. F2CLIB\libf77\abort.c' ....it is "no return value",
however, some of routines were compiled without any problem.
Seems to me that for some particular routines , I used wrong switches (options).
I followed the README file in CLAPACK package, but
the option MSDOS and/or NON_UNIX_STDIO did not solve compilation problems.

Any suggestions ?

Does exist some kind of step by step procedure for compiling the CLAPACK via
Watcom C/C++ 10.6 compiler under win95 ?

Are CLAPACK pre-compiled libraries available?

Thanks a lot

Slavomir MATAS e-mail : matas@saske.sk
IEP SAS
043 53 Kosice
Slovak Republic


By Hauke Hildebrandt on Tuesday, October 27, 1998 - 12:06 pm:

Id like to use Lapack++ (a real Newbie) but have problems to get it running. Which libraries must I link (for safety Ive linked all that were made/used during compilation of CLapack and Lapack++)? And: When trying to compile a program that uses Lapack++ I get loads of undefined references, e.g. in function ilaenv_ to s_copy. This may have something to do with C++ and F2C, but I dont know how to resolve the problem. Can someone help me? Thanks.

Hauke


By Lih-Sin The on Wednesday, November 11, 1998 - 04:03 pm:

Hi,
i try to build lapackpp libraries in DEC Alpha running OSF 4.0A with egcs-1.1b compiler. However,
i stuck with this error messages:

g++ -I../include -DALPHA -c lautil.cc
In file included from lautil.cc:29:
../include/lautil.h:17: warning: ANSI C++ forbids declaration `' with no type
../include/lautil.h:17: abstract declarator `int' used as declaration
lautil.cc: In function `int LaEnvBlockSize(const char *, const class LaGenMatDouble &)':
lautil.cc:43: warning: implicit declaration of function `int ilaenv_(...)'
*** Exit 1

if i use the gcc 2.7.2.1 compiler in Intel Linux,
it compiles without a problem.

i'll appreciate very much if you could tell me
how to solve this problem or find a way around it.

has anyone compile lapackpp with egcs compiler in
DEC Alpha OSF? can i get it?

Thansk,
lihsin

lihsin@astro.phys.clemson.edu


By Mesfin on Tuesday, December 1, 1998 - 04:57 pm:

Hi, I have been trying to use the lapackpp package for a unix work station. But, I get the following message at the end when I do the "make all" testing after down loading the libraries:

Unresolved:
s_wsfe
do_fioi4
e_wsfe
do_fio
f_exit
f77fatal
collect2: ld returned 1 exit status
*** Error code 1 (bu21)
*** Error code 1 (bu21)

can any one help me how to fix this.


By Lu on Tuesday, January 19, 1999 - 05:09 pm:

HI, there,

May I have a question.

I am trying to solve a system of highly nonlinear
equtions. But using linear search method, it always converge to a LOCAL minimum point instead of the GLOBAL mimimum point. From most books in this field, the only way to search the exact solution is trying another initial guess, but I have no idea about this so far.

My question is you know where and how i can get the fortran subroutine to search the GLOABL minimum points.

Thanks a lot in advance

Yours sincerely.

w.m.lu


By Bill_allen9 on Thursday, February 18, 1999 - 03:57 am:

Hi, does anyone know the algorithm used by
passport readers?

Thanks!
bill_allen9@hotmail.com


By Cpenghua on Wednesday, March 17, 1999 - 02:06 am:

Any one out there has a version of SVD
in C code for COMPLEX input matrices?
Most of the versions that I come across
are for real input matrices!
Thanks


By Manuel Capouet on Monday, March 22, 1999 - 07:11 am:

Hi,

I m using the parallel version of lapack, scalapack. I wanted to solve a 1000 by 1000 dense non symetric matrix using PVM on 9 processors but the software returned a message error saying there was a lack of memory (Total Memory required is about 4 Mbytes). Since my available total memory is quite larger than that, I suppose it is an artificial limitation of the soft.

How could I change it ?

Thanks

manuel Capouet
mjcapoue@vub.ac.be


By Ingemar_Hulthage on Thursday, April 8, 1999 - 07:03 pm:

I can't find the routines s_cat, s_cmp and s_copy, needed in CLAPack. I can guess at what they do, but I prefer to download them.
Where are they ?

Thanks,


By Ingemar Hulthage on Friday, April 9, 1999 - 08:00 pm:

S_cat, s_cmp and s_copy found in libf2c.


By Whzh on Thursday, May 13, 1999 - 11:27 pm:

Does anyone know how perform mathematical operation to 3 slaves with the use of 1 master and get the final output by getting processing time?


By Paul Chuang on Friday, July 23, 1999 - 12:56 am:

Does anyone know how to use the sobol sequence
in Numerical Recipes C? Here I want to generate the normal(0,1) using quasi random numbers:
/*********************************************
float *x;
int M=3,n1=(-1);

x=vector(1,M);
sobseq(&n1,x);

for(i=1;i<1024;i++){
sobseq(&M,x); quasi_ran1 = x[1];
sobseq(&M,x); quasi_ran2 = x[1];

normal_ran
=sqrt(-2*log(quasi_ran1))*sin(2*pi*quasi_ran2);

printf("%10.6f ",normal_ran);
}
/**********************************************/
It seems what I got are not normal(0,1).
Can anyone help me?

Thanks a lot.


By Julian Ho on Thursday, August 5, 1999 - 09:04 am:

I'm currently using MS VC++ and have been using
a pre-compiled .dll of CLAPACK from phc.chalmers.se/~agris. But I have now
discovered errors in this compilation. Is there anywhere else a version of CLAPACK for VC++ can be downloaded or any tips on its compilation.

Any help is very much appreciated.


By Marco DAttilio on Monday, September 6, 1999 - 10:43 am:

Hi,
I have this problem:

I have large sparse matrix and I need to solve a problem of the type: x'=Ax+bu
whit A square sparse matrix, I need of a C routine to calculate the inverse of A and multiplied a sparse matrix for a vector.
Could you help me?

my Email:
fontana70@hotmail.com

Thank you!


By Alfred J. Svobodnik on Wednesday, September 29, 1999 - 02:01 am:

Hi everybody,

I am looking for a parallel version to solve symmetric complex matrices. Currently I am using zspsv (zsptrf/zsptrs) from LAPACK. Are there any suggestions for me?

Many thanks

Alfred


By Santoso on Wednesday, October 27, 1999 - 07:28 am:

Does anyone would like to share his/her experienced in using LSODI from ODEPACK.
1. Is it just CALL LSODI from main program ?
2. Is there hidden trick or hint to use it ?
3. How to get subroutine xerrwv ?

Thanks a lot


By Marco Agus on Wednesday, January 19, 2000 - 10:46 am:

Does anibody know how to solve the following problem?
I need to compile a simple program in Microsoft VC++ but the linker gives me the error:
LAPACK.lib : fatal error LNK1106: invalid file or disk full: cannot seek to 0x374def12
Error executing link.exe.
Does anybody know the reason?


By Tom Bell on Wednesday, January 19, 2000 - 03:41 pm:

I have a polynomial with real coefficients. The order is anywhere from 5-9. I'd like to solve for the roots as quickly as possible. However, I'm only interested in real roots that fall within a predetermined interval. Though a general root finder would work, I was wondering if there was an alogorithm out there that, by taking into account my interest in only a subset of the polynomial's roots, would run faster.

Please cc to tombell@integrinautics.com. Thanks in advance.

Tom


By Magus74 on Thursday, January 20, 2000 - 03:41 am:

Hi folks,
I have this error in Microsoft VC++5.0 when I use
LAPACK.lib:
LAPACK.lib : fatal error LNK1106: invalid file or disk full: cannot seek to 0x374def12
Error executing link.exe.
Does anybody know how to solve it?


By Colin Zhao on Thursday, January 20, 2000 - 01:53 pm:

The .lib might have been built with another version of MSVC.


By Sridharan on Sunday, January 23, 2000 - 04:22 am:

I have a function f of one variable T
(running from 0 to 1) that is only
continuous. There is no simple expression
for the function. However, given a value
of T, I can compute value of f by an algorithm.
I want to find the smallest root of
f (smallest positive T for which
f(T) = 0).
I have done a search on Netlib and I have
not been able to locate any program that
finds the smallest/first root. Any
pointers/comments would be appreciated.
Thanks. Please reply to sridhara@postmark.net.


By Sridharan on Wednesday, January 26, 2000 - 07:47 am:

Hello -
This is with reference to my posting on
Sunday, January 23, 2000 with regard
to finding the first root of a function
that is only continuous. I had
inadvertantly put a period after net
in my e-mail address. The correct
address is : sridhara@postmark.net
Apologies for any bounced mails.

I have tried some variations of the
bisection method and the regula falsi
method and have not had much success.
It is not clear whether I should pursue
approaches based on random walks to find
roughly a point where f may be negative
and then backtrack. Thanks. Sridharan


By Dag on Sunday, March 12, 2000 - 08:26 am:

Morning everybody,
I have a problem with linear programming. There
is a linear optimization problem that seems to
be cyclic and the procedures I have used up to now
failed to give a satisfactory answer (they always
find a trivial solution).

Does anybody know any procedures that takes into
account the degeneracy of the problem?

Is the degeneracy a common problem? I am asking for it seems to me that I've found a huge set of
degenerate problems.


By Jim on Tuesday, March 21, 2000 - 09:34 am:

Hello everybody!

I'm a desperate newbie and i need to use lapack++ with Visual C++ 6.0. I tried to build the library with Vis.Stud. but nothing, i then tried with DJGPP but i can't overcome the problems i have. Can anyone give me a step by step instruction? Has any of you used the library on a pc with Visual C++?

thank you in advance
jim


By Jim on Tuesday, March 21, 2000 - 09:35 am:

Hello everybody!

I'm a desperate newbie and i need to use lapack++ with Visual C++ 6.0. I tried to build the library with Vis.Stud. but nothing, i then tried with DJGPP but i can't overcome the problems i have. Can anyone give me a step by step instruction? Has any of you used the library on a pc with Visual C++?

thank you in advance
jim


NOTE: posting new messages to this forum has been disabled due to security problems.

Add a Message


This is a public posting area. If you do not have an account, enter your full name into the "Username" box and leave the "Password" box empty. Your e-mail address is optional.
Username:  
Password:
E-mail:
Post as "Anonymous"