R, OPR, etc

[Skyehawk]

Hello to the rather specific part of the FIRST community I am targeting,
I was reading over OPR calculation and other such calculations in order to calculate offensive contributions and some ideas started to weasel its way into my head, therefore I must pursue them. I should pick up the R language relatively easily, as it seems to share many common elements with Python and Matlab. I was wondering are there some resources that you guys could point me towards to aid with my learning? I am slowly falling deeper end deeper into the statistical analysis rabbit hole and any help to ease the future confusion I may have is greatly appreciated.

Happy New Year!
Skye Leake

[Oblarg]

This textbook is oriented towards psychology students, but is a good introductory textbook for just about anyone:

https://health.adelaide.edu.au/psych.../teaching/lsr/

Be sure to download Rstudio, it's about a billion times easier than using R through the standard shell.

[Skyehawk]

Originally Posted by Oblarg

Be sure to download Rstudio, it's about a billion times easier than using R through the standard shell.

One of my housemates went through an entire semester of Biometry working straight from the shell, it made my skin crawl. A proper GUI is a must and was installed mere seconds after I installed R.

[EDIT]
I have been using this website as an introduction. It is geared more towards novice programmers but seems to provide strong overviews of the various components of Rstudio.

[Ether]

https://www.chiefdelphi.com/forums/s...7&postcount=24

[Skyehawk]

Originally Posted by Ether

https://www.chiefdelphi.com/forums/s...7&postcount=24

The plan is to slowly work to that. I have a lot of learning to do. I have bookmarked that thread for later use, thanks for linking it in though.

I have some concerns about my package palette. I am sure I will find what I need in time through trial and error; but is there anything additional packages that may prove generally useful that are basic 'need-to-haves'?

Thanks,
Skye Leake

[Oblarg]

Originally Posted by Skyehawk

The plan is to slowly work to that. I have a lot of learning to do. I have bookmarked that thread for later use, thanks for linking it in though.

I have some concerns about my package palette. I am sure I will find what I need in time through trial and error; but is there anything additional packages that may prove generally useful that are basic 'need-to-haves'?

Thanks,
Skye Leake

ggplot2, because it's far easier to use than the default plotting package once you learn it, and reshape2 are the two I'd say are "need-to-haves."

[Ether]

Originally Posted by Skyehawk

...I have a lot of learning to do. I have bookmarked that thread for later use, thanks for linking it in though.

Originally Posted by Skyehawk

I was reading over OPR calculation and other such calculations in order to calculate offensive contributions and some ideas started to weasel its way into my head, therefore I must pursue them... I was wondering are there some resources that you guys could point me towards to aid with my learning?

If you need any links to posts and threads here on CD that explain the math underlying the computation of OPR let me know.

[sirwin]

I'm not sure if you've seen the other thread that addresses R and OPR, but I wrote an R package for downloading and manipulating FRC data in R. You might find it useful to peruse the source code, which is on github. I also have links to the R resources that I used in the Getting Started Guide for the package. Links to everything are at http://irwinsnet.github.io

sirwin

[Ether]

Originally Posted by sirwin

I'm not sure if you've seen the other thread that addresses R and OPR... Links to everything are at http://irwinsnet.github.io

Great resource Stacy. Thanks for posting that.

FWIW, I linked to your post about that several days ago, and the OP indicated he was aware of it.

For others who may not have seen it yet, there is a short paper here that shows how to use sparse matrix technology in R to speed up OPR computation by a factor of 10 or more.

[Citrus Dad]

Originally Posted by Skyehawk

I am slowly falling deeper end deeper into the statistical analysis rabbit hole and any help to ease the future confusion I may have is greatly appreciated.

Fall into that rabbit hole! Learning R will have much bigger payoffs down the road. Many of the best statisticians I know use it and it can be a tremendous work skill to have.