ten computer codes that transformed science
In a “Feature” article of 21 January 2021, Nature goes over a poll on “software tools that have had a big impact on the world of science”. Among those,
the Fortran compiler (1957), which is one of the first symbolic languages, developed by IBM. This is the first computer language I learned (in 1982) and one of the two (with SAS) I ever coded on punch cards for the massive computers of INSEE. I quickly and enthusiastically switched to Pascal (and the Apple IIe) the year after and despite an attempt at moving to C, I alas kept the Pascal programming style in my subsequent C codes (until I gave up in the early 2000’s!). Moving to R full time, even though I had been using Splus since a Unix version was produced. Interestingly, a later survey of Nature readers put R at the top of the list of what should have been included!, incidentally including Monte Carlo algorithms into the list (and I did not vote in that poll!),
the fast Fourier transform (1965), co-introduced by John Tukey, but which I never ever used (or at least knowingly!),
arXiv (1991), which was started as an emailed preprint list by Paul Ginsparg at Los Alamos, getting the current name by 1998, and where I only started publishing (or arXiving) in 2007, perhaps because it then sounded difficult to submit a preprint there, perhaps because having a worldwide preprint server sounded more like bother (esp. since we had then to publish our preprints on the local servers) than revolution, perhaps because of a vague worry of being overtaken by others… Anyway, I now see arXiv as the primary outlet for publishing papers, with the possible added features of arXiv-backed journals and Peer Community validations,
the IPython Notebook (2011), by Fernando Pérez, which started by 259 lines of Python code, and turned into Jupyter in 2014. I know nothing about this, but I can relate to the relevance of the project when thinking about Rmarkdown, which I find more and more to be a great way to work on collaborative projects and to teach. And for producing reproducible research. (I do remember writing once a paper in Sweave, but not which one…!)
Xi'an's Og 
April 24, 2021 at 10:51 am
Before computers became an inanimate object, they were mostly women with 10-key adding machines and tables. At Los Alamos Lab for the Manhattan Project they stoked the boilers of the Bomb in support of history’s largest sustained in-person collaboration of Nobel prize winning physicists.
One of them may have been working a problem and needed the square of 48 and wondered aloud. Hans Berthe would pipe up “2,304–rule of squares near 50.” Someone else would need the cube root of 1,728, and Feynman volunteered 12–cubic inches in a cubic foot of water.
Now imagine that by some improbable alternate universe that Berthe and Feynman were there ONLY to do math parlor tricks. Would Robert Oppenheimer have traded them in on an IBM 701 and FORTRAN?
Not a chance.
Science thinking is done in real time. Writing, card punching, job batching and debugging was not a REPL exercise. Trains of thought derail while waiting.
For a box of HP-35 RPN calculators and supply of batteries? THAT would be a no-brainer.
For science as it is actually done, the humble pocket calculator was the biggest boost to the practice of science in the 20th century.
April 23, 2021 at 4:26 pm
I agree with the readers that Matlab should be high on the list. And I find it hard to believe that you have never used the FFT :-) Spectral estimation, or just visualizing (a numerical approximation of) the Fourier transform of a signal…
April 23, 2021 at 5:36 pm
Not knowingly, at least…
April 23, 2021 at 8:17 am
[…] article was first published on R – Xi'an's Og, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here) […]