Slashdot Mirror

R

'nuff said. Oh wait, bit more to be said to pass the lameness filter. Ok, and I need to type a couple more characters per line. Wheew, that's fun! Type type type.

I think you correctly identify the strengths of NR, and if I do remember correctly (it's a long time since I read the whole anti-NR page), they give them some credit for it too.

The main problem with NR is that as long as you understand the methods there fully, and realize their limitations, you're on safe grounds. Sometimes you can appreciate them fully from NR itself, and sometimes you can't, and a few of the methods are just flawed. So, you can use the stuff in NR when you fully understand it, but once you've reached that level, you've probably allready outgrown it... :-)

The reason why I learned OOP is pretty much that I had to. I'm at the University of Oslo in Norway, and OOP was invented here and implemented in SIMULA. So they really forced us to use SIMULA and program OO, but in just a couple of basic courses, so I didn't a prolonged forced exposure.

Examples where OO concepts are clear include SIMULA and the S system (my favorite implementation R). SIMULA is a full-blown OOP system (but there are various reasons why it failed), while S has just a few OO features, but the features that have been implemented are easy to understand, they incorporate some essential concepts, and they are very powerful.

Then, you have C++ which is also a full featured language, but where the concepts are not that clear and easy to grasp.

When OO gets muddy is usually when it is built on the top of existing languages. Perl OO is a bit muddy, but not too bad, but if you look at e.g. IDL it gets rather bad IMHO.

I haven't looked at NAG, but you're not getting all the source code? Actually, I find it weird that one does rely on closed source in science, when science depends on the full disclosure of all relevant material. Well, the computations done is arguably the most important aspect of most analysis done today, so all source code should be available for public scrutiny.

I would recommend R for many applications. It has some lightweight OO that is very efficient, it's a really beautiful language. It is also very rigorous, you will find the best things in Numerical Recipes there, but there is also a lot of code that has been through formal peer review.

Many of my spreadsheets qualify as "spreadsheet applications" and need quite a bit of flow control type logic.

Then you might want to drop the spreadsheet entirely, and go for R. R is a very beautiful system for statistical computing and graphics, and it is very powerful. I know there is quite some R code in Gnumeric, most of the statistics comes from there.

I think there has been some talk about a more extensive interface between Gnumeric and R, but I don't know what happened.

There is some more high-level GUI tools for R as well, but I have never had any use for them, so I haven't even compiled it, but there might be some things that are sufficiently spreadsheet-like there that you can use.

Statistics
R-Project
Solid software, similar to Splus with possible linkage to C, C++ and Fortran.

Image Analysis
UTHSCSA Image Tool
Functional image analysis with script language.

Fortran
VFort
Stadnard MDI app and g77 compiler. Good environment for inexperienced Fortran programmers.

R comes with Woody (Next Debian release).

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.

Here are some other grey areas:

1) installing modified GPL code on a machine for a customer to evaluate -- RMS has told me that if the customer controls the machine, it is distribution, but if the code modifier controls the machine, it is not.

Perhaps the problem is that you want to keep secrets? That's perfectly respectable, but that's not what this license is all about.

Quite seriously, the GPL does impose some serious burdens on distributors. Fortunately, these can all be overcome by distributing only in source code form. Configure and make are slick enough that this is entirely practical for me, and I'm certainly no programmer. I can't set this stuff up myself, but I have no trouble using it to install big projects like R .

I plotted the number of CDs shipped along with a linear fit (using R) and put it in a postscript file (I'm not going to have it there forever).

It is quite clear, it is a linear trend, with quite a lot of variation. I'll do a brief Analysis of Variance on the linear fit in a seperate reply (damn, why can't I have the PRE-element?).

There is a lot of sucky OOP support out there. IDL is really bad. R on the other hand, has some very simple OOP support, but it is certainly very nice, and makes the job done very nicely. I haven't really gotten the hang of the Perl OOP stuff, but then, my Perl stuff isn't really sophiscated.

R is a clone of the statistics programming language/environment S-plus, which was invented at Bell Labs a number of years ago. It's a very nice, interpreted language with elegant support for vectors, functional programming, and even some basic object-orientation.

Grab the source from your nearest CRAN mirror or the Debian package.

R is a clone of the statistics programming language/environment S-plus, which was invented at Bell Labs a number of years ago. It's a very nice, interpreted language with elegant support for vectors, functional programming, and even some basic object-orientation.

Grab the source from your nearest CRAN mirror or the Debian package.

Don't forget R.

After this, and after seeing what some closed source packages does, I really don't have much faith left. Mathematica, however, seems good, I only wish I could dissect the source.

It can be used on Unices (the use of emacs as frontend is strongly recommended, it's an excellent collection of modes for it in the ESS package) and on Win32, there's a GUI for it on that platform.

Then, there's the PDL, offering "Number crunching capabilities for perl", I haven't used that either, but I hear it is good. Probably meant as an alternitive to IDL, that I have used, I didn't like it's syntax, so I did some hacking to replace the features I needed from IDL in R.

As an alternative to MATLAB, I hear SCILAB can be used. It doesn't seem to evolve very fast, and I haven't used it.

Now, I miss an alternative to Mathematica. Mathematica is the only proprietary software (and that I would think about using...) I can think of that still is better than Open Source alternatives. I have heard that some emacs package does have symbolic mathematical capabilities, but I don't what it is.

There is a slightly different perspective that I would like to point out. This would allow more Statistics departments to switch over to Linux. The major (and superior) development environment for Statisticians is S. A Linux port for S already exist. More importantly there is R, a free implementation of S which is equally good if not better than the commercial version. So, now you have essentially everything you need as a statistician on Linux.