R programming

>>58425011
For statistics, it is. Source: My statistics prof is an R fanboy and contribooted his own packages and has his own CRAN repo.

Also, it is pretty easy to fuck up your data in excel t be h senpai.

That said, as a regular programming language it sucks.

>>58425138
Would just use it (atleast for now) for that statistics course. Excel sounds to me like a complete fuckup. Though as little as we do in statistics it is probably okay to explain everything, but I still think it is shit.

>>58425232
>Excel sounds to me like a complete fuckup. Though as little as we do in statistics it is probably okay to explain everything, but I still think it is shit.

For statistics it sure is. Remember, the british austerity program was based on a miscalculation done by Excel.
Also: There are so many ways to fuck up CSV data imports.

>>58425702
>british austerity program was based on a miscalculation done by Excel
holy shit I didn't know that

>it's actually true

>>58425833
it's actually true?

damn, i remembered that so vaguely it could have been wrong

>>58425861
https://www.theguardian.com/politics/2013/apr/18/uncovered-error-george-osborne-austerity
https://www.bloomberg.com/news/articles/2013-04-18/faq-reinhart-rogoff-and-the-excel-error-that-changed-history

>projected -0.1% growth for economies with +90% public debt
>without the error it would have been +2.2%

I thought you were taking the piss of Excel but I googled it, unbelievable.

>>58425861
I think you're referring to the Reinhart and Rogoff paper about austerity that was overturned after an excel off-by-one column error was found. The article justifies austerity policies, but it's doubtful that British austerity policies were directly based on one research paper.

That said, you can make stupid mistakes anywhere. But you can't unittest spreadsheets...

>>58425886
>I thought you were taking the piss of Excel but I googled it, unbelievable.
It's not so unbelievable though, shit like that happens every day even to scientists, so why would a bunch of retarded politicians advisors know better.

All those implicit conversion thingies without warning are against all interface design and engineering rules and I'm curious how many lives that has claimed already.

>>58425011
R is superior for Statistics... But your professor might be smart by teaching you in excel because that is very common to use in the job.

But nothing beats R for statistics, the proprietary shit does not even come close.

>>58425886
>I thought you were taking the piss of Excel but I googled it, unbelievable.

Easy to make mistakes no matter what tool you use. Which is why the sciences have stuff like peer review. It's not only to catch charlatans.

>>58426089
If your second point were true, peer review would require data disclosure and code review. It doesn't, so it can't catch that.

Peer review these days is an exercise in acceptability politics for the journal. It's disgusting.

its p much just a fancy command line calculator

R's packages are just overwhelmingly convenient. Once you've gotten even somewhat used to R, your overall workflow should go much faster than the drag the box down fest that is Excel.

I'd say that the biggest problems with Excel have to do with it's frequent rounding errors, and how prone it is to misuse. You wouldn't believe how much the big 4 accounting firms have fucked up because they thought that Excel's numerical solvers could replace well known algebraic methods derived using MLE. I almost found it embarrassing to look at some of their scripts.

>>58426127
>If your second point were true, peer review would require data disclosure and code review. It doesn't, so it can't catch that.

The only peer review I am experienced with is chemistry. And there ARE detailed information available for those things. Maybe not data disclosure, but complete methodology disclosure so that you can do your own tests with the same experiments.

Why computer fucks don't share code is stupid when the research is the code. So how do they do review and test the research?

>>58425011
>Is R a good replacement for Excel?
Is a spoon a good replacement for a shovel?

Why don't you first figure out the difference between a statistical language and a spreadsheet program

They should have a class on it