Free Scientific Software for Developing World?

FlashBoltzmann asks: "I'm a physics student in the US working with a group of physicists, mostly from Africa, who are interested in helping their colleagues on the continent obtain free software for scientific and educational use. Often, many science departments in Africa have little or almost nonexistent funding to purchase new software packages, especially for scientific research or education. Some know of the free software available but say it takes up large amounts of time over often slow internet connections to find and obtain it. I am asking for any recommendations on freeware or open source software, for any operating system, that anyone knows about. We are looking at the Debian version of Linux for a lot of the great software that comes with it but resources for MS Windows would be helpful as well."

"Free educational software of any level is appreciated though we prefer college and graduate level software. Also, field specific software is great, e.g. software for condensed matter physics. Eventually we'll probably combine the software on CDs to be distributed to these scientists. Any help is appreciated especially with programs that perform simulations, mathematical and statistical analysis and plotting, compilers, lab software, etc. The users of the software will most likely be physicists or mathematicians."

Comments

The one to begin with...

[mirko]

...is TeX.

This typesetting program was originally aimed at the scientist. I don't know of any other software that produces nicer documents.

Re:The one to begin with...

[zuvembi]

TeX doth indeed rock. I started using it for (of all things) doing a family cookbook. Which I found it did many times better than a normal word processor.

Now I've started using it for reports and documents at work (anything to save me from MS word). I started using Lyx at first, but I found I actually liked doing LaTeX by hand better.

An old math favorite

[imrdkl]

Was Are you ready for Calc III. This, and alot more math software can be had from the UofA Math Software Page.

Open Channel Software

[Xpilot]

Well there's always this.

I recall /. running a story about NASA donating some stuff to this site...

Reasons to not bother with Windows

[wrinkledshirt]

I know this isn't entirely the point of your question (and more than a little bit of this is motivated by my anger towards MS's recent settlement), but I just thought I'd put forward the idea that you shouldn't bother with Windows at all.

If you're hurting for cash for software, the outlook is probably not all that great for hardware too, right? The cutting edge of Linux and the various BSDs all run well on hardware that the latest Windows versions cough and sputter on.

Also, providing a Windows learning environment is only going to encourage use of Windows down the line, which will require further investments, software AND hardware upgrades etc.

If you're working with a blank slate, and these people need training anyway, might as well put it towards something that won't come back and make serious demands on your checkbook. Save the money for additional learning resources, a CD burner to replicate software for yourself (this is legal with the Linux and BSD OSes), etc. Don't go down the proprietary road, or else soon enough you'll be dealing with the same MS-driven crap the entire Western world is trying to handle right now.

Re:Reasons to not bother with Windows

[Reality Master 101]

Anyway, back to my Red Cross story, they wanted me to go through these two Excel spreadsheets of volunteers to weed out duplicates and it looked like a dauntingly tedious process,

Of course, you could have just merged the two lists in Excel, sorted them, and then written a VB program to delete the duplicates. What, 2 minutes work?

Just because you don't know how to use a tool doesn't mean it's a bad tool.

Re:Reasons to not bother with Windows

[Remote]

they wanted me to go through these two Excel spreadsheets of volunteers to weed out duplicates and it looked like a dauntingly tedious process, so I converted the spreadsheets to comma-delimited text files and booted the Dell I was given to work with to a Linuxcare CD, mounted the disk, and used cat and grep to go through the list at several times the speed I was expected to go at. If I were a "Perl Monk" I am sure that I could have blown through it in even less time.

Or you could have done the same with tools written for Windows, only replacing type for cat and loosing the CD boot/mount thing.

Of course, I see your point. Unix is inherently more flexible when it comes to doing non-standard things, and utilities talk to each other much more seamlessly. What I want to note is that there is an enourmous world under Windows beyond what Microsoft offers, some of it open source. If a staff is already comfortable with an OS, I don't think it would be wise to migrate, except if there are at least two people proficient with Unix (which might well be the case) and time and budget for training. OTOH, maybe your example sprung up to my eyes just because I'm so used to using Borland's grep that I'm beginning to feel it's part of the OS!

A note on distributions: I imagine the situation in Africa as being similar to the one in Brazil as far as hardware is concerned, that means a lot of cheap/old cards that may not be supported by many distros. If you (the submiter of the original question) face that kind of problem, try Conectiva Linux. They have paid special attention to that aspect.

Try NA/European physicists & mathematicians

[hearingaid]

I worked for a little while in a government research library, and about half the people in the building were both scientists and programmers. They developed a lot of their own tools, and most of them were coding for some *nix, many on Linux.

They didn't care about other people getting their code. I would expect universities to be the same way.

As for bandwidth, that's much less of a problem now with CD burners. I'm assuming your Third World people have CD-ROMs, but given that, if you can talk to some First World scientists & get them to burn and ship, it might well be cheaper.

SAL: Scientific Applications on Linux

[gylle]

SAL is a good resource for finding science apps that run on Linux. Worldwide mirrors, many apps are free.

SAL - Scientific Applications on Linux

[Anonymous Coward]

This is most likely the most complete site out there when it comes to science on linux. http://sal.kachinatech.com/

Netlib and more

[apsmith]

The standard resource for free scientific software (unfortunately mostly written in Fortran) is Jack Dongarra's netlib: http://www.netlib.org/

It's best in linear algebra (matrix problems etc) but there's other good stuff in there - FFT routines, statistical stuff, some deep mathematics, and more... Also, not free, but good, is the standby Numerical Recipes book, which includes source code for a large variety of uses, particularly solution of nonlinear optimization problems.

Other stuff is available free from the supercomputer centers - at least they used to give stuff away free, though NCSA at least seems to have tried to make money off their things lately...

Re:Netlib and more

[Wile E. Heresiarch]

I do quite a bit of number crunching. Here are
some of the resources I use:

Netlib (www.netlib.org) -- Yes, it's mostly Fortran, but that's a good thing! Just use f2c (easy to find) and translate to C if that's what you want. Don't underestimate the power of decades-old programs -- old == widely used and well-tested.

StatLib (lib.stat.cmu.edu) -- Collection of statistical software, in various languages, including C, Fortran, and S.

SAL, Scientific Applications on Linux (sal.kachinatech.com) -- a very large collection of links.

Freshmeat (www.freshmeat.net) -- Not scientifically oriented, but there is much scientific stuff there, along with all kinds of miscellany.

Octave (www.octave.org) -- A package for matrix manipulations, similar to Matlab, but free. Useful for all kinds of problems.

R (www.r-project.org) -- An implementation of the S language for statistics, but also useful for general problems, similar to Octave. S+ is a commercial implementation of S.

Well, that ought to be enough to get started. To echo something other posters have mentioned -- don't even bother with Windows software. If your budget is tight, save your money for hardware, don't waste it on the MS tax.

Re:Netlib and more

[altstadt]

Why is it that whenever this topic comes up, everybody always talks about Octave and gnuplot as if these are the only things available?

If you want a self contained program that runs on multiple platforms, take a look at Euler. This is about as close as it gets to being a MatLab clone.

astro software & some stats too.

[fiiz]

It is actually not that easy to find free physics software.

For professional astronomy software, I recommend http://star-www.rl.ac.uk/

Some nice but steep stats software in the R project http://www.r-project.org/

And you can use Octave & gnuplot for basic maths. (admittedly not as good as mathematica,matlab or some other maths package.)

This URL http://www.seul.org/sci/seul-sci10.html has a review of linux & GPL packages that are useful to scientists.

It is also probably worth asking some of the software vendors if they would like to donate something, as really, you never know! (if the cause is good...)

Good luck!

fz

Physics Analysis Workstation.

[matthayes]

Developed at CERN
Great for graphical representation, and statistics. Released under GPL.

I remember using it about three years ago under Red Hat for reconstruting cosmic ray showers. Can't see any possible problems with Debian...
It was great for what I was doing.

Matt.

Re:Physics Analysis Workstation.

[ErfC]

I've used PAW quite a bit for my work. It's quite effective at stuff, but the documentation is abyssmal. I found it really really painful to learn, and as a result I didn't get very far into it.

Something almost as hard to learn but somewhat easier to actually use is Physica, developed at TRIUMF. It's the main program I used to do my M.Sc. analysis work. :)

For data aquisition (and generally running an experiment), I strongly suggest looking into MIDAS. It's really powerful, and has a web interface (with optional password protection), electronic log book, etc, which is really helpful for experimenters to keep tabs on things from home. Especially when "home" is in another city (or even country).

University of Tokyo Physics

[r.suzuka]

I am a student of Physics at the University of Tokyo and my superiors have instituted a plan that accomplishes many of these aims.

We have correspondence programs with several universities in Africa in which we will provide to them our outmoded hardware. It is unfortunate that often, we are not able to replace our hardware as often as we would prefer, but when we do, we attempt to find a physics department without adequate hardware.

Also in the course of completing their theses, graduate students must write various software tools to assist them. The copyright to these tools belongs to the University if I am not mistaken. Although my University does not distribute these freely as some would prefer, they are sometimes provided to the other universities which have the hardware necessary to run them (with the consent of the programmer student of course).

Yes, there is more that may be done, but I believe that we are working to genuinely assist other physics programs which are less fortunate that we are in some respects. Does anyone else know of similar programs?

R. Suzuka

Theorem proving software

[reachinmark]

For the more logic inclined mathematicians, as well as anyone interested in structured proofs, theorem proving, etc, the HOL theorem prover is a very powerful engine. On the down side - it can be tricky to learn, but there is a large quantit of documentation and a big user base to help get people started.

See: http://archive.comlab.ox.ac.uk/formal-methods/hol. html amongst other pages.

Stuff that I use

[Mochatsubo]

Python with Numeric Python and Scipy make for a fine numerical computing environment (www.python.org, www.pfdubois.com/numpy/, www.scipy.org).

The GNU scientific library (GSL) can be found here: http://sources.redhat.com/gsl/

Intel Image Processing Library (C): http://developer.intel.com/software/products/perfl ib/ipl/

Intel Signal Processing Library (C): http://developer.intel.com/software/products/perfl ib/spl/

VTK is an *extensive* visualization toolkit (C++): http://public.kitware.com/VTK/

So Internet is too slow to download distros?

[arminh1974]

Now here's something where the average joe without coding skills can help promote free software. How about offering to burn distros & RPMs/DEBs and mail them to africa or other places where the infrastructure isn't so great? The costs shouldn't be too horrible ... maybe we can set up a network of volunteers for something like that? I myself don't have a CD-burner at the moment (relocated from US->EU recently), but I can punch out a simple database-driven website quickly ... if anyone's interested, mail arminh(AT)usa.net ... maybe we can get something going?

Electronics schematics capture package

[Grab]

If you're in need of a package to draw electronic circuits and do general schematic capture stuff, check out Pulsonix. These guys have made a fully-featured electronics schematic capture program freely available to anyone who wants to download it. The only downsides are (a) it's 19MB, and (b) it's Windows-only. Highly recommended.

The package also contains a PCB design package and other good stuff; however, you have to pay to unlock these (not excessive amounts by the standards of most design packages, but £1-2K is a lot for someone in a developing country or your typical hobbyist). If you're on a real budget, just use the schematic capture part to produce netlists, and then use an old DOS/Win3.11 PCB layout program such as BoardMaker for the PCB design.

Grab.

Re:Electronics schematics capture package

[Grab]

Agreed. But (a) you don't have to be honest :-), and (b) you are getting a seriously good bit of kit for the exchange. Since this is the only freely-available (as in "pay no money") schematic capture program I know of, there ain't much choice.

Grab.

Re:Quick question

[duffbeer703]

Microsoft products, Windows included, is licensed for signifigantly cheaper prices abroad.

Microsoft also donates alot of Windows licenses and other software to scientific and other institutions overseas.

useful research programs ( and libraries )

[hagbard5235]

Here are the things that I am aware of that have been found quite useful:

For graphing:

• Grace ( formerly xmgr )

For Numerical Analysis:

• GSL ( Gnu Scientific Library ):

language bindings for perl,python, and C++ for GSL are also available.

Check out the Scientific Computing FAQ: which I've been having trouble reaching so you might want to try the Google cache of it.

Re:A better solution

[DarkZero]

Some people have a problem with abandoning their country. Besides... if you wanted a video game from Japan, or a beer from Canada, would you go and live there so you could have it? No, you wouldn't. You'd import it, and that's exactly what they want to do.

It's people like this that make us Americans look arrogant and stupid.

GAP -- Computational Group Theory and more

[stevelinton]

GAP is a powerful software system for computational abstract algebra and discrete mathematics, especially group theory. See http://www.gap-system.org for details (including mirrors) and download. It's distributed under a "copyleft" not too different from the GPL.

If you want to use GAP for research or teaching and can't download it (we've had people whose bandwidth is too low, and people whose countries do not allow arbitrary internet downloads for political/religious reasons) let us know (mail one of the addresses on the Web site) and we can usually manage to send a CD.

Steve Linton

Seriously... check out Freshmeat.net

[eno2001]

They have a scientific/engineering Visualization section that has a lot of cool stuff. Here are some examples:

K-3D modeling, rendering and animation software (Win32 as well):
http://midas.psi.ch/

Isotopic Pattern Calculator (Link may be wrapped):
http://www.uni-duesseldorf.de/MathNat/pc1-AK_Wei nk auf/ipc/ipc_d.html

MayaVi (Visualization Software):
http://mayavi.sourceforge.net/

MIDAS (Data acq software for particle physics):
http://midas.psi.ch/

GraphThing (Graph Theory tool):
http://members.optushome.com.au/davidsymonds/gt/

GNU TeXmacs (Technical writing tool, great for technical docs with formuli):
http://www.texmacs.org/

There are 130 projects on Freshmeat, which is probably just the "tip of the iceberg".

I am not a troll. ;P

Re:A better solution

[mehfu]

Well, I think these countries are in need of scientists. They need people with higher education to develop a more stable economy. Why not instead put all american scientists (with all their funding) in countries with less opportunities to give money to research? This would give a lot of job opportunities as well.

Your idea is probably the most stupid I've seen so far today.

Re:Physics Analysis Workstation - ROOT

[Dave Bailey]

Also try the ROOT package. It's also developed at CERN (by the PAW people) but is in C++ (with a built in C++ interpreter) and has much more to it than PAW.

It's aimed at the Particle Physics community but is currently in use in a wide range of fields from Astronomy to banking!

Oh yes, runs on Linux and Windows...

Re:A better solution

[Huge Pi Removal]

I have to disagree with that...

It would be really nice if research could be done, freely, across the world. If everyone leapt into America every time they wanted to {start a business|do some research|etc, etc} then America would be:
1. Overcrowded, and
2. Resented by the rest of the world for "stealing" all its talent.

I don't see that Africa is "cut off" from the rest of the world. They have phone lines to send e-mail down, and geographically Africa's a lot nearer to the rest of us in Europe than America is. Or do you think "cut off" simply means "cut off from America"? And anyway, if no-one else biult up their research facilities, that means all research ends up in America anyway, which is a Bad Thing(tm) (see above paragraph).

Personally I'd love to see these people being helped to do what they want, in a country that they feel is home. I sure as hell'd think twice before going to some states in the US if I were black.

Hope that wasn't an anti-American rant. I know my posts usually are...

Octave

[s20451]

Octave is a matrix manipulation package, released under GPL - basically a clone of MATLAB. It has scripting capabilities, which allow development of simulation software.

TISEAN

[Skorpion]

A very good package for chaos-theory-oriented numerical data analysis is TISEAN.

It does excellent job on its part. There is also some documentation on the site, including one of the creators' Ph. D. thesis that explains some of the theory behind the software. On Linux it requires gcc and GNU Fortran complier to compile (compilation is pretty straightforward).

I also found GNU awk extremely useful at numerical data analysis. You also would want to include Python and NumPy - python extension for numerical computations.

HTH

Alex

Some useful free tools

[spike_gran]

I managed to get my Physics PhD using almost entirely free tools.
The thesis was written in LaTeX, using emacs, and made printer friendly with dvips.
The data plots were done in gnuplot.
The simulations were written in c with gcc or Fortran with g77. For the matrix analysis algorithms I used LAPACK. For minimization routines I used some of the Numerical Recipes routines, which aren't free software exactly, but Numerical Recipes is an easy book to buy used off Amazon.
I know that all of this stuff is really old-skool, but, it all works fine.

Things I use at work...

[doop]

I'm doing a PhD on simulations of soft condensed matter, and mainly use either free software, or stuff we wrote in-house. Off the top of my head:

• VTKis a very good package for scientific visualization.
• Maxima is a Free computer algebra system, a bit like Mathematica. It can solve equations, do calculus, plot things, produce TeX output of what you've done, and lots more. Incredibly useful for long tedious bits of algebra.
• gnuplot is a versatile graphing package (2D and 3D, but maxima or VTK are IMO better for 3d stuff). As well as graphing, it can try to fit arbitrary functions to your experimental data.
• LaTeX -- it's very hard indeed to typeset equations better than LaTeX can.

If you're interested in condensed matter physics (or a few other areas), then you should have a look at the Los Alamos E-print server, which contains preprints of a lot of scientific papers.

Re:Do we want advanced scientists working overseas

[the eric conspiracy]

I say keep the third world in the dark ages.

So we will live in perpetual fear of madmen like Bin Laden?

Until we have enlightenment on a global basis, we are all in danger of falling from civilization.

octave and matlab (student version)

[adapt]

while you can get Octave for free with most Linux dists. and you can contact the Octave people for tips about installation. the US student version of Matlab is cheap and does not have matrix size limitations AND there is a Linux version in the package. also, if you check Matlab website [mathworks.com , mathtools.net] , they have a huge ftp site of free goodies, i.e. Matlab toolboxes, that probably can be used with Octave too.

since i installed my student version of Matlab at home, i have used less my Octave. Matlab also can be bought at academic prices, which are still too expensive for cash-strapped academia.

as for linux vs. windows, if you have to leave you computer on for 10 days for a simulation, then linux stability is a nice bonus...

lotsa *nix sci freeware in french

[rainTown]

I found these links for linux scientific freeware on this page http://www.freepatents.org/liberty/logiciels.html

Its in French... but then again the majority of my African friends speak it.... there is a lot in there .....sorry for the lack of form....and i didn't check all the links.... hope its useful...

Sciences et ingénierie
Scientific Applications on Linux http://SAL.KachinaTech.COM/
Index très complet d'applications scientifiques et professionnelles (gratuites, shareware ou commerciales) qui tournent sous Linux.

Statistiques
fiasco http://www.fsf.org/software/fiasco/index.html
xldlas http://a42.com/~thor/xldlas/
MacAnova http://www.stat.umn.edu/~gary/macanova/macanova.ho me.html
R http://www.ci.tuwien.ac.at/R/contents.html
Calcul formel
MuPAD http://www.mupad.de
Maple http://www.maplesoft.com/
Mathematica http://www.wolfram.com/
Macsyma http://www.macsyma.com/
Magma http://www.maths.usyd.edu.au :8000/u/magma/
Macaulay2 http://www.math.uiuc.edu/Macaulay2/
Singular http://www.mathematik.uni-kl.de/~zca/Singular/
Analyse numérique
Scilab http://www-rocq.inria.fr/scilab/
Matlab http://www.mathworks.com/products/matlab/mlover.sh tml
Octave http://www.che.wisc.edu/octave/

Here's an idea

[Flakeloaf]

Why not get the African countries to prosecute Microsoft for anticompetitive practices and force MS to just give them the software to pay penalties?

windows trap

[adapt]

yes, MS sells campus licenses at EXTREME discounts (like $20 for Office and less for the OSs), but the hardware requirements are heavier.

most people in academia are not swimming in cash, so this means old hardware, and an array of diverse machines connected to a server. linux is the ideal software partner for a small research group, in my opinion.

the other factor, as somebody else pointed above, is that GNU or public tools are used by almost everybody. most papers are swapped in .ps format and written in TeX on (insert favourite editor here :). in windows you can do the same, but integrated tools like Scientific Workplace cost money, and they are not really needed after you learn some shell and vi tricks.

still, the crucial factor that made me wipe out windows for linux was stability. when you do not have a double Xeon crunching numbers, you appreciate the fact that linux will not crash during the 3 days it must be ON. ;-)

Re:Do we want advanced scientists working overseas

[onion2k]

Terrorists exist because they come from uncivilized, barbaric nations

Like the Unabomber you mean?

Re:bsd ports

[tollieman]

Any means of obtaining free software IMNHO is not pain-ass.

Let's see, start a download and go to sleep while it finishes, or...

Pay out of your ass (alot more painful for most).

An analogy...

Go to a strip club, and spend tons of money sticking dollars wherever, or...

Rent a Pr0n video, watch it, throw money at the TV, when it's over, pick up the money...

same thing.

linux is the answer

[dummkopf]

being a physicist myself (theory/computational physics) i have noticed that the main trend is to get rid of the expensive sun workstations and geat cheap pcs with linux on them. while we can have endless fights of what distro is the best, it seems that (at least in america) redhat (7.2 is highly recommendable and available via cd) is the choice for most scientific groups.

not only is it a free os, it also provides ALL the core tools you need to do research! for example you have TeX (+ several excellent text editors), the whole gnu compiler suite (and debuggers), excellent plotting tools for data and image manipulation (gnuplot, gimp, xgrace, ...) and many more. Institutions like CERN, or space telescope provide full packages with tools to analyze all kinds of data.

there are a lot of other scientific applications you can get for free for linux if you are in an academic environment and which are awesome tools to use for researchers. i have seen many responses already with good pointers to different places (SAL, freshmeat, CERN, IBM Open DX).

finally, once can also make computational clusters with linux -- really inexpensive ones!

mupad

[platypus]

Take al ook at mupad

It's some sort of mathematica lookalike, superior in some cases and they have free versions.

It's been a while since I used it, but it was great.

Re:mupad

[dido]

Have tried this. My ripes include the fact that is that it is somewhat slow (compared to proprietary computer algebra systems like Maple or Mathematica anyhow), and (gasp) it uses XView. Yes, the old GUI toolkit Sun once created for Open Look and their OpenWindows desktop.

The worst part about it is that it's NOT open source/free software, which means that you're basically betting that the folks at the University of Paderborn who developed it aren't going to stop maintaining it or will suddenly stop making versions of it that are freely usable. Keep this in mind if you decide to use it.

GIS system

[Kris Warkentin]

Grass is a very powerful, free GIS system which is quite useful to scientists. A good GIS application can be used for any number of things such as terrain and weather modeling, migration pattern tracking, etc.

Re:Do we want advanced scientists working overseas

[duffbeer703]

I was referring to Mideast terrorists in particular. They are a particular breed of terrorist movement and cannot be compared directly to terrorist organizations in Spain or Ireland. Unfortunately, I cannot edit my previous comment and clarify it.

The difference is that Mideast nations are RUN by the uncivilized, barbaric dictators and petty kings.

Western democratic governments are accountable to the will of the people. While organizations like the CIA have supported rotten regimes in Central America and Asia, that support rarely lasts a long time, since administrations and political power shifts every few years.

IRAF for Astronomy/Image Processing

[andrew cooke]

http://iraf.noao.edu/iraf-homepage.html is the standard data processing package in American/British astronomy (and possibly Europe too these days). I just noticed it is packaged inside Debian...

Although aimed at astronomy, it would be useful general image processing (particularly good at automating procedures over many images).

Fortran compilers

[the gnat]

This isn't going to help them as far as the bandwidth problem goes, but Intel now offers their Fortran compiler free for unsupported noncommercial use. This includes F90, which opens up a lot more opportunities. It does need a (free, still) license, so it's a little tricky to obtain, but still very worthwhile. As far as I've seen it's the only free F90 compiler for Linux and the only free F77 compiler besides g77, and it's likely to be far faster than g77 as well.

R (aka GNU S)

[KjetilK]

For my thesis in astrophysics, I have almost exclusively used the R-system. I find it brilliant. It was developed for statistic, but IMHO, it can be used for any numerical computational task, though in some areas, it may need more development (for example, it lacks 2D FFT, but that should be easy to fix.

R comes with Woody (Next Debian release).

Re:Quick question

[SirSlud]

[offtopic]What's a non-terrible ask slashdot, or are you just upset that half of them don't involve games anymore? Given current events, I find it hard to believe that trying to get some of the more poor places in the world up 'n running in so far as marketable skills and data goes constitutes a poor Ask Slashdot.[/offtopic]

Connections and Connectivity

[Joe Decker]

A note to FlashBoltzmann (admittedly a bit of a digression)--I notice that you point out that a lot of these folks have slow connections. I suspect that's true in a direct sense, but isn't it true that continental bandwidth to/from Africa is still pretty limited as well? It seems to me that there's a chance that an archive of related software, located on the African contient, might help the downloading time issues and give you a place to put a bunch of related software to make it easier for folks to find....

Numerical Python

[Devil's Avocado]

One of the issues I've often run up against when doing scientific programming is the desire for a *real* programming language to support the number crunching. This often caused huge frustration for me when I used Matlab and IDL. One of the nicest solutions I've found for numerical programming is the Numerical Python package. (http://www.pfdubois.com/numpy/ ) You get the numerical expressiveness of Matlab or IDL with the power of Python as a programming language for the half of your program that *doesn't* deal with crunching numbers. (In my experience it's actually usually more than half, even in heavily numeric code!)

Here are a few more links:
The Python website: http://www.python.org
The Scientific Python Project: http://www.scipy.org

Cheers,
-DA

A few favourites we use

[Professor J Frink]

In the past few years I converted our lab over to Linux and here are some of the tools we use for analysis:

• GCC for C/C++/FORTRAN coding. It's free, it's not the fastest in the world but it's competent.
• Octave is a great, free replacement for Matlab. For general data manipulation it seems fine, where it really lacks relative to Matlab is in the GUI.
• Gnuplot is a great all-round, all-purpose, scriptable plotting tool that can also do fitting. For general everyday tasks gnuplot gets used a lot in our lab.
• SciGraphica is a great 2d/3d/vector/polar/ plotting and analysis package. It is a little like an Origin clone so is pretty easy to pick up, and can be extended with Python plugins. I am one of the developers ;0) (although far too busy atm to contribute, anyone want to help?). More suitable for publication-quality plots and still heavily in development. A new release is imminent. Plug ;0).
• teTeX is the main (La)TeX distribution for Linux and you'll most probably have it in Debian anyway but for writing reports, articles, books, theses, even letters you shouldn't need to use anything else. Really.
• OpenOffice if you have to deal with mad, crazy, annoying .doc using people.

There's plenty more where they came from. Most distrbutions come with a lot of these things anyway. These are mainly analysis or document tools, there's plenty of other things for both these areas and any other which plenty of other posters have shown. I've written a little guide for my local group. Some of it's out of date (and some of it's wrong but I have better things to fix) but it does have a list of common tools we use. And, of course, SAL is a pretty comprehensive database of unix tools. HTH.

Re:and the second is

[kilgore_47]

I think we should follow star trek's prime directive; when we encounter less advanced cultures, we should refuse to give them our technology and insist that they discover it on their own.