This is for IE7 to hold div open
Click on a topic to go to that section.
The IMSL library contains over one thousand Fortran subroutines and functions. These routines provide easy access to high quality implementations of numerical methods in mathematics and statistics.
Mathematical routines include: linear algebra, eigensystem analysis, interpolation and approximation, integration and differentiation, solving differential equations, Fourier and Laplace transforms, finding solutions of nonlinear equations, optimization, and special functions.
Statistical routines include: regression correlation, analysis of variance, categorical and discrete data analysis, nonparametric statistics, tests of goodness of fit and randomness, time series analysis and forecasting, covariance structures and factor analysis, discriminant analysis, cluster analysis, sampling, survival analysis, life testing, and reliability, multidimensional scaling, density and hazard estimation, probability distribution functions and inverses, and random number generation.
The IMSL subroutines are called like ordinary Fortran subroutines, then the IMSL libraries are linked into the executable when the calling program is loaded. As discussed below, the method of linking depends on the system you are using.
IMSL is proprietary software and cannot be transferred to a machine that is not licensed to run it.
The IMSL Fortran Numerical libraries are now included as part of Visual Numerics' Computational Technology Toolkit (CTT). IMSL Fortran Numerical libraries V5.0 comes as part of CTT 6.0 and is available on the ITS Unix servers running on the Solaris operating system: ccwf.cc.utexas.edu and uts.cc.utexas.edu. If you want to use IMSL on your departmental machines then you must obtain a license for them. The University of Texas does not have a source license for the IMSL library. If you are interested in obtaining either a source license or a binary license contact:
Visual Numerics Corporate Headquarters 1300 W. Sam Houston Suite 150 Houston, Texas USA Tel: 800-364-8880 713-784-3131
They can also be reached by e-mail at info@houston.vni.com
source /usr/local/imsl/CTT6.0/ctt/bin/cttsetup.cshThe command above can be put at the end of your .cshrc (or .profile) so that it will be executed automatically for you. It is important that you issue this command only after you have set your environment variables such as your PATH and MANPATH. Once you have executed the script, use:
f90( or f77) [ compiler options ] $FFLAGS $LINK_FNL filename
to compile and load the program in filename using the default (shared)
version of the IMSL library. It is also possible to explicitly specify
the use of the static or shared library. To load the static or shared
versions of the IMSL Fortran numeric libraries, respectively, use:
f90( or f77) [ compiler options ] $FFLAGS $LINK_FNL_STATIC
filename
or
f90( or f77) [ compiler options ] $FFLAGS $LINK_FNL_SHARED
filename
To generate an executable eitest that uses the shared version of the
IMSL libraries, compile the example file eitest.f with:
f90( or f77) -o eitest $FFLAGS $LINK_FNL eitest.f
Logical Points to
imslieee_share The IEEE shared library linker options file
YELLOW$DRA5:[IMSL30.ADTVNI.LIB.AXPVMS]IMSLSHRE.OPT
imslieee_static The IEEE static library linker options file
YELLOW$DRA5:[IMSL30.ADTVNI.LIB.AXPVMS]IMSLSTCE.OPT
imslibg_share The G_FLOAT shared library linker options file
YELLOW$DRA5:[IMSL30.ADTVNI.LIB.AXPVMS]IMSLSHRG.OPT
imslibg_static The G_FLOAT static library linker options file
YELLOW$DRA5:[IMSL30.ADTVNI.LIB.AXPVMS]IMSLSTCG.OPT
imslpsect The program section linker options file
YELLOW$DRA5:[IMSL30.ADTVNI.LIB.AXPVMS]IMSLPSECT.OPT
For a detailed discussion of the contents of this
file see chapter 3 of the OpenVMS Linker Utility
Manual.
These logicals are used to compile Fortran programs with the desired
version of the IMSL Library. The /opt switch identifies them as options
files to the linker. To compile the test program eitest.f
using the shared IEEE floating point library use the command:
fortran/float=ieee eitest.f
link/nomap eitest.obj, imslieee_share/opt, imslpsect/opt
The following commands will compile and link eitest.f using
the static IEEE floating point library:
fortran/float=ieee eitest.f
link/nomap eitest.obj, imslieee_static/opt, imslpsect/opt
The commands below compile eitest.f with the shared G_FLOAT
floating point library:
fortran/float=g_float eitest.f
link/nomap eitest.obj, imslibg_share/opt, imslpsect/opt
The static G_FLOAT floating point library is used with the commands:
fortran/float=g_float eitest.f
link/nomap eitest.obj, imslibg_static/opt, imslpsect/opt
Note: The IMSL Library is designed to allow quantities to underflow
gracefully and be treated as zero. However, due to a flaw in the DEC
runtime libraries, floating underflow warning messages may be generated.
DEC has been made aware of the problem and will fix it in a future release
of the libraries. Users should expect and not be concerned by these
messages.An example program
You may be using a completely different machine or operating system, but
this example will provide some guidance even in that case. The following
program uses the IMSL routines EI and UMACH.
C-----------------------------------------------------------------------
C
C Purpose: Call the IMSL routine EI to evaluate the exponential integral
C for arguments greater than zero and the Cauchy principal value
C for arguments less than zero.
C
C Declare variables
INTEGER NOUT
REAL EI, VALUE, X
EXTERNAL EI, UMACH
C Compute
X = 1.15
VALUE = EI(X)
C Print the results
CALL UMACH (2, NOUT)
WRITE (NOUT,99999) X, VALUE
C EI( 1.150) = 2.304
99999 FORMAT (' EI(', F6.3, ') = ', F6.3)
END
These are pointers to information on the Internet that you may find helpful. Be aware, though, that because these resources are maintained at remote sites, the information may not always apply to you, and may not always be available.
SSC provides consulting services for a wide range of applications by electronic mail, telephone, and office visits. For an overview of the consulting services available to you, see SSC Consulting Services.
Assistance for IMSL is provided by SSC Mathematical Support, who you can reach by electronic mail at:
math@ssc.utexas.edu
Please include a description of the problem including any particular IMSL routines you might be trying to use, a log of the relevant parts of your compilation or execution, and any data your program might be using. Also, if possible include a campus or other local telephone number where a consultant can reach you. You can also request an appointment if you feel it would be better to talk directly with the consultant.
Additional information about this or other mathematical and statisticalsoftware can be obtained from the SSC Consulting home page at http://ssc.utexas.edu/consulting/index.html.