Archive for the ‘Computing’ Category

Competent versus exceptional

January 3, 2015

Here is Paul Graham on the need to allow immigration of programmers into USA:

We have the potential to ensure that the US remains a technology superpower just by letting in a few thousand great programmers a year. What a colossal mistake it would be to let that opportunity slip. It could easily be the defining mistake this generation of American politicians later become famous for. And unlike other potential mistakes on that scale, it costs nothing to fix.

So please, get on with it.

Of course the entire discussion is based on the premise that

… there is a huge variation in ability between competent programmers and exceptional ones, and while you can train people to be competent, you can’t train them to be exceptional. Exceptional programmers have an aptitude for and interest in programming that is not merely the product of training.

I do not know how far it is true and if true, why it is so. Of course, that would be the most interesting question to ask and the answer might be worth the pursuit!

Simulation methods and their theoretical basis

November 6, 2013

Simulation methods are useful only if there is a solid theoretical basis for whatever work is being done; if there is not, they are just ‘miserable tricks’  …

Pierleoni in his personal tribute to Giovanni here.

A book and a open sourceware

August 4, 2013

I am about to complete Michel Danino’s The lost river (thanks to the recommendation of a colleague — who strongly recommended the book by passing his copy on to me). I enjoyed reading it a lot. The book has piqued my interest on Indus civilization. The text-book material that we read in schools is so inadequate. I would love to read an authoritative and up to date account on all aspects of this civilization.

Another colleague from IIT-Hyderabad has suggested that I try elmer; I have loaded it on my comp; looks very interesting and I would love to teach it sometime to my students!

Stop complaining: start killing

July 24, 2013

The lousy patents, says Joel. A good post.

Trello

May 1, 2013

Just heard about Trello from Joel and signed up! Looks promising; may be I can use it to manage things around in the lab and office! Let us see.

The first commandment of computing and some academic put-down

September 4, 2012

Via John  Hawks, here is a link to Titus Brown’s thoughts on anecdotal science and the need for sharing codes:

Closer to home, I think I can attribute some of my collaborators’ impatience with me to this attitude of mine. I want to do good, solid, robust computational science, as well as relevant biology; my schtick is, at least at the moment,computational methods. Since my collaborators tend not to be computationally focused, they don’t always get the point of all the computational work. Some of them are either more patient or more relaxed about the whole thing — if you’re wondering why Jim Tiedje is co-authoring papers on probabilistic de Bruijn graphs, well, that’s why :). Some of them are less patient, and it’s why I would never recommend a bioinformatics analysis position to anyone — it leads to computational science driven by biologists, which is often something we call “bad science”.

What’s the bottom line? Publish your methods, which include your source code and your parameters, and discuss your controls and evaluation in detail. Otherwise, you’re doing anecdotal science.

Also via John Hawks, I got the link to this academic put down (Pullum on Chomsky):

For example, he claims that children have an innate grasp of the psychic continuity of persons, and his basis for the claim is that his grandchildren enjoy a story in which a baby donkey gets turned into a rock. It is surprising to see a point this feeble published even once, but Chomsky has now put it into print at least three times.

McGilvray invites Chomsky to explain how “Merge” brought humanity the gift of arithmetic.

The attempted exposition is a train wreck. Chomsky confuses the empty set with zero, binary operations with unary ones, functions with the sets on which they are defined, the natural numbers with their set-theoretic construction, and both with the theory of arithmetic. McGilvray drinks it all in – and appends a commentary note that unfortunately elaborates the second of Paul Benacerraf’s two celebrated non-equivalent set-theoretic reconstructions of the natural numbers, when Chomsky was struggling to outline the first. It is embarrassing – like overhearing a conversation between two undergraduates about a mathematics lecture that neither has understood.

Following the 141 pages of transcribed chunks of conversation are some lengthy appendices and commentary notes, mostly just pedestrian restatements of Chomsky’s increasingly eccentric linguistic doctrines, and together with the glossary, bibliography and index making up 56 per cent of the whole book.

At least half a dozen “interview” volumes of Chomsky’s unrehearsed musings have been published since Mitsou Ronat created the genre in 1977. This one is McGilvray’s fourth book-length homage to Chomsky by my count. He is entitled to his view that anything Chomsky can be induced to say should be typed up and distributed. But why are university presses publishing stuff like this, devoid of carefully framed ideas, results or scientific data about language?

Presumably the guaranteed sales from having Chomsky’s name on the cover are too tempting to resist. Buyers should beware.

As John Hawks has titled his post, this is chomping of Chomsky!

Parallel computing: a good resource

August 24, 2012

After CUDA, now I am getting interested in plain old parallel computing (mpi and stuff like that) as well as in openCL (thanks to a collaborator, Prof. Phanikumar). So, I found this webpage, especially, this piece to be very helpful.

The computing pleasures

July 23, 2012

I recently upgraded my desktop by enhancing the RAM (from 4 to 16GB)  and adding a GPU (GeForce GTX580) card. However, I had trouble in using all the RAM; it would show only 4G, and I learnt that I either have to update the kernel or load a 64bit version. I decided to go for the second option and, while at it, decided to load the Precise Pangolin (Ubuntu 12.04).

My initial attempts to load Pangolin failed because I did not turn on the noapic, nolapic and nomodeset options while loading Ubuntu. Once I got that right, rest has been a sail. I really, really like the feel of the Unity desktop even though I am still learning the tricks. On the whole a nice experience so far.

Now, I am able to get a simulation with 4096×4096 size running (which takes only half of the available memory)! This probably is the first time I am running such a large scale simulation; the previous big simulations used to stop at one quarter of this size (at 2048×2048) and I am very, very excited!

Next step is to figure out CUDA and start using the GPU for computations, and needless to say, I am looking forward to it.

Scientific/engineering/mathematics software

July 22, 2012

A good list and contains several of my favourites: FFTW, GSL, Sage, Octave and R; scilab and gnuplot are missing though.

Optimizing software and not getting wet in the rain

July 21, 2012

A rather detailed manual on software optimization (C++) and other optimization resources; via Andrew at Statistical Modeling, Causal inference, and Social Science.

A link to a paper in European Journal of Physics as to whether you should walk or run in a rain — in Bombay, where, you get drenched even if you are under the umbrella, it might be a good idea to learn to enjoy the warm, big fat drops hitting you!