I prefer excel.

I use R. I recommend you use Rstudio IDE. You can input the data in R or in excel, then make an output of the data using the plot function.

>>54984375
R code is so much easier to reproduce though. I just grab some old code, edit it here and there, read in a different data set, boom. Excel is icky mouse clicking.

I love R for basic scripting and plotting.

If you really dig down into the theory, it's actually kind of a shitty language, but if all you're doing is simple shit you'll never notice.

Also don't use RStudio, the response lag in the REPL is horrendous and the extra features aren't worth it.

i'd like to use R but i read that it has certain subtleties like ways of making things faster that aren't what someone picks up from programming regular loops or something. i use python, numpy and matplotlib for all that stuff. it's super easy too. by now i know how to open a file and plot it from memory

>>54987209
there are some rules of thumb, but it's an interpreted language, any differences in speed caused by not using idiomatic expressions is negligible

what you're referring to is the fact that most native functions are actually calling compiled code that runs a loop over a vector under the hood. you CAN always loop over a vector yourself in R (and there are some rare cases where that's actually the only easy way to do something), but usually it's faster to just pass a function a vector and let the compiled code do the work for you without all of R's overhead

but like i said, for most applications of R the difference is negligible

Can you apply custom styles to plots(a la css)? The stock appearance looks straight out of 1995.

>>54987351
i'm not sure if it's exactly what you'd call custom styles, but yes, you can customize pretty much everything about the plots and make htem look however you want.

there's also some very nice third-party packages (ggplot2 is one of hte most popular) that produce much nicer out of box images

>>54987343
that makes sense, given that i'd be mostly going to be working with vectors and matrices if i'm using it. i hardly use loops when i'm doing numerical stuff with numpy anyway

>>54985280
>If you really dig down into the theory, it's actually kind of a shitty language

Howso? R is on my "to learn" list.

>>54987343
For many applications of large dataset processing, it's easy enough to farm out the processing into R servers.

>>54987413
http://r.cs.purdue.edu/pub/ecoop12.pdf is a good read

tl;dr the OO implementation isn't great, no real concurrency support except for wizard-level users, and the way R works under hte hood is really inefficient and woudl require a massive overhaul to fix

it doesnt affect the end-user in the slighest except that it's not as efficient as it theoretically could be