Luke PembletonWhat’s your #rstats #ROrigin story? I started my PhD in 2011 and initially was using Excel macros 😱 to perform optimal genetic combination selection, taking over all the office PCs at night for pseudo parallel processing. Within a few months I attended an internal 1day quantitative breeding course which used R. I was immediately hooked 🤩 everything felt so natural and I rewrote all my analyses the next day in R. 2years later I successfully submitted my first pkg to CRAN, & the rest is history 🥳
George Stagg (Moved)@lwpembleton In 2008 I started a Mathematics and Statistics degree at Newcastle University. Several modules had computational aspects, where we were introduced to a whole bunch of languages: R, Python, Matlab, Maple, even Fortran! R was the language that clicked for me the most, I loved how quick and simple it is relative to the others to get up and running with numerical simulation and visualisation. Later RStudio came along and made it even easier once it was my turn to teach R.
Ken Butler has moved@gws @lwpembleton I am older than you: we learned Minitab, and yes, Fortran for numerical analysis.
@nxskok @lwpembleton I believe there were still some modules teaching with Minitab when I left, but I never had the pleasure myself. I actually like Fortran a lot, it’s ace for high performance parallel codes using MPI on HPC. My PhD was based around writing a numerical solver in Fortran for a certain class of PDEs used for physical simulations of superfluid turbulence.
@gws @lwpembleton not at the same level, but I was definitely coding things like five-point Runge-Kutta in Fortran in those days.
Jonathan Carroll@nxskok @gws @lwpembleton 5?!? I only coded a fourth-order to solve the equation of state for 'neutron' stars for my PhD 😃
(please do not critique my code, I was never really trained to write that)
@nxskok @gws @lwpembleton I don't think I can dig out any of the F77 code I worked on, but I did convert a lot of that to F90. As an undergrad 'intern' I played with some F66 weather prediction code from the US Gov but yes, that was all before my time. My supervisor had stories of sending his punch cards via the guy on a bike.
@jonocarroll @gws @lwpembleton a stack of punched cards carefully secured by rubber bands, no doubt. I used punched cards very briefly: submit the cards, *wait*, eventually get the cards back plus (hopefully) some line-printed output in your "pigeonhole", fix as many bugs as you could find, repeat.