home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!wupost!crcnis1.unl.edu!moe.ksu.ksu.edu!zaphod.mps.ohio-state.edu!darwin.sura.net!spool.mu.edu!olivea!charnel!sifon!isaac
- From: isaac@gildor.Physics.McGill.CA (Martin-D. Lacasse)
- Newsgroups: sci.math.num-analysis
- Subject: C-based fitting and data manipulation program
- Message-ID: <1992Nov17.003301.9355@sifon.cc.mcgill.ca>
- Date: 17 Nov 92 00:33:01 GMT
- Sender: news@sifon.cc.mcgill.ca
- Reply-To: isaac@physics.mcgill.ca
- Organization: Center for the Physics of Materials, McGill University
- Lines: 63
- Nntp-Posting-Host: gildor.physics.mcgill.ca
-
-
- If you use GNUPLOT and do day to day fitting (linear or non-linear)
- or Fourier transforms, column manipulation and plotting (and
- moreover, you hate awk) then this might interest you.
-
- Hi everyone!
-
- In course of my (student) work, I had to deal with a *lot* of data
- files to fit, modify, plot, Fourier transform, etc. This lead me
- (over the last year) to write a mathematical front end to GNUPLOT.
- At first, this program was built for a particular personal need,
- but very fast it spreaded in our research group and I soon had to
- answer requests for further developments.
-
- Basically, the program runs with GNUPLOT (version 3.2 is fine,
- earlier versions need to be recompiled to prevent exit on error)
- without any modification. It allows the user to use the syntax of
- C-shell for having aliases, while and foreach loops, Macros can
- also be defined. This is how I dealt with the multiple files to
- transform/fit/plot. This is particularly useful when one has to fit
- the parameters obtained from a series of previous fits. For the
- data (column) manipulation, the program uses a C-interpreter I
- wrote, with some particular adaptations. It supports variables,
- vectors (the data columns) as well as strings. All vector
- assignments are implicit loops, Thus, commands like: x=1; X=x++
- define a vector X ranging from 1 to the vector size. Most (if not
- all) C math library functions are available for data manipulation.
- So, commands like Y = sin(X*pi)^2 assigns vector Y to the square of
- the sin() of PI times each element of vector X. With these
- extensions, the resulting flavour of C becomes a nice data
- manipulation language.
-
- As it stands now, it is only supported on UNIX and compiles
- on AIX, ULTRIX, IRIX, and HPUX.
-
- It is fully documented, and a series of past applications are
- provided as examples. If you know C and C-shell programming, you'll
- feel at home. If not, you'll find that using it is relatively easy.
- Although the program is still under development, it is in a
- relatively stable state.
-
- Some of the included mathematical routines were based on the ones
- found in Numerical Recipes. Most of them were deeply modified to
- allow user-defined application, dynamic memory, and elegant error
- recovery. However, before I know more about the fuzzy (and
- sometimes funny) limits of numerical method copyrights I suggest
- that you don't use that program unless you already have access to
- NR routines. If someone can enlighten me about this point, please do!
- (Use e-mail, I don't think this group is the good place for that).
-
- This program is called FUDGIT and is available by anonymous ftp at
- ftp.physics.mcgill.ca.
-
- Enjoy!
-
- ---
-
- Martin-D. Lacasse internet: isaac@physics.mcgill.ca
- Physics Department voice: (514) 398-7027
- McGill University fax: (514) 398-8434
- Montreal
- Quebec, Canada
- &&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
-