Webcast: Actuarial Analytics in R

[ultrafilter]

http://www.revolutionanalytics.com/n...nalytics-in-r/

Thought this might be of general interest.

[Colonel Smoothie]

Some people in our department (me included) have signed up for it.

[Expunge]

I watched it live...

I'm interested, but now trying to do anything on my own, I'm completely frustrated. I now need to be a learn R class, not just a demonstration of what it can do.

After playing with R briefly, and search the web for tutorials etc, it seems to be a simple language that has a steep learning curve. Most recently I've been programming in objective-C for iOS in Xcode, and R has a long way to go, to become more friendly, (or maybe that is a huge compliment to Apple)

Maybe another difficulty is that I'm rusty on my statistics, and need to go refresh those skills first.

Does anybody have a great getting started source for this. (I'd rather find something free online as an absolute beginner before buying a book to learn more)

[ultrafilter]

Unfortunately there's not a lot of intro R stuff out there that's aimed at people who haven't done some programming before. R Twotorials looks pretty promising, but it's hard for me to judge the quality, given that I've been programming for a while now.

Edit: I missed that you have some programming experience. R's a little different from some other languages. I've heard very good things about The Art of R Programming, and I like the bits I've seen.

[KSBurke]

The best way to learn R is to look at some code and figure out what it means and then make alterations to it.

I found these books to be helpful
A Handbook of Statistical Analysis Using R
Extending the Linear Model with R

As far as free stuff goes, take a look at
SimpleR-Using R for Introductory Statistics
R Graph Gallery (Type out and run the code.)
Statistics Using R with Biological Examples

Learn how to subset data, aggregate and summarize data. Import and export using read.table, write.table and related methods.

[Sub_Zero]

It would be cool if one of these new programming training websites had some R content.

(udacity, codeacademy)

[ultrafilter]

Originally Posted by KSBurke

The best way to learn R is to look at some code and figure out what it means and then make alterations to it.

I found these books to be helpful
A Handbook of Statistical Analysis Using R
Extending the Linear Model with R

As far as free stuff goes, take a look at
SimpleR-Using R for Introductory Statistics
R Graph Gallery (Type out and run the code.)
Statistics Using R with Biological Examples

Learn how to subset data, aggregate and summarize data. Import and export using read.table, write.table and related methods.

There's also Hadley Wickham's Essential R Vocabulary.

[diegol]

I agree that R has a steep learning curve. When I was learning it, I found it frustrating that even to import sample tables before getting to the real thing I had to write code. I yearned for a simple interface to interact with other apps, and there is one in the R Commander package. I wish I had known this at the early beginning.

[Fish Actuary]

For my new job I am having to work extensively with R as my boss is obsessed with it and thinks that it can handle all of our modeling and data analysis needs (this is for conservation genetics, rather than actuarial science). I've basically just jumped in head first with a background in using fortran, matlab, and SAS. If time isn't horribly crunched and there's a push from above to use it, I suggest just jumping in head first into trying to do whatever your task is. You can google a lot of what you're trying to do by searching for R and then a few keywords.

Another thing to look at is Springer Use R book series which may have a book or two covering your topics

[Fish Actuary]

Originally Posted by diegol

I agree that R has a steep learning curve. When I was learning it, I found it frustrating that even to import sample tables before getting to the real thing I had to write code. I yearned for a simple interface to interact with other apps, and there is one in the R Commander package. I wish I had known this at the early beginning.

R Studio is also fairly nice.