Permalink
Switch branches/tags
Nothing to show
Find file
Fetching contributors…
Cannot retrieve contributors at this time
91 lines (81 sloc) 3.15 KB

Useful Links

https://bitbucket.org/tavisrudd/r_users_group_1/src https://bitbucket.org/tavisrudd/r_users_group_1/get/tip.zip https://bitbucket.org/tavisrudd/r_users_group_1/get/tip.tar.gz

Introductions And A Quick Survey

Name and background

Experience with R?

Statistical training/experience?

Programming experience?

If no R experience, what have you heard about R?

If no R experience, what do you want to learn?

Thanks to our Sponsors!

Pulse Energy, our host

Revolution Analytics

Goals

Today

stoke your interest!

have some fun playing with it

line up speakers for next meeting in May

For the Group

Not just about R!

Data management, analysis and visualization in general

meetings every 2 months

mix of interactive sessions and talks

R as an Environment

long history (S, S-PLUS)

huge, rapidly growing community

thousands of libraries http://crantastic.org/ http://cran.r-project.org/ Lots of interesting blogs: http://www.r-bloggers.com/ http://learnr.wordpress.com/

optimized for fast, INTERACTIVE data exploration

awesome help system

very easy to learn the basics, while doing useful things

A Few Quick Examples

employment data

unemployment

US crimes by state

murder_vs_burglary

timeseries lattice graphics of sales data

lattice_example

Crayola colours over the years

crayola

calendar heat map of blog posts

heatmap

R as a Programming Language

assignment with <- instead of =

not zero indexed, starts at 1..

vector-based rather than scalar based (very important!)

functional, in a loose sense

higher order functions

lexical scoping

an “expression language”

looping in R syntax is inefficient

… but there are higher order functions that make it very efficient and easy.

not side-effect free, like Haskell or Erlang!

Very useful built-in data types (similar to Python)

with a very useful vector/array/matrix slicing syntax

functions take keyword arguments, with optional defaults

“object-oriented”, but with “generic functions”

obj$attr NOT obj.attr

this.is.just.a.variable.name this_is_a_valid_varname for modern versions of R, but wasn’t in the past

a bit inconsistent and idiosyncratic, but awesome nonetheless

Exercises

import a small tabular / numeric dataset into R

Preferably your own data from real life projects, but there is plenty of interesting CSV data on the web if you can’t use your own.

print summary descriptive statistics of the data

examine the data structures that R stores it in

create subsets of it

create same basic charts of the data

if applicable, explore relationships in the data (correlation, etc.)

Resources

http://www.statmethods.net/index.html

http://www.r-tutor.com/

What’s next???

A source of meeting / talk ideas:

http://www.meetup.com/R-Users/#past