Embedded Files
One of the great features about R is that you can write up detailed code to perform pretty much any mathematical/statistical function that you can imagine. Such as calculating an Anova.....
AArrrrrrrgh!!! Hell no!
"I am NOT going to re-invent the wheel, I will borrow" ~ motto of an experienced R user.
There is an extensive list of "Packages", which are basically beautiful gifts to you. A package consists of functions that perform some mathematical or statistical job on your data, and return you summary results. In Tinn R look for the symbol that looks like a "tiny package wrapped in a bow".
R packages that I use frequently:
NORTEST - 5 different ways to test your data for normality (which you should do before applying hypothesis testing stats to your data):
- Anderson Darling
- Cramer Von Mises
- Lilliefor's (Kolmogorov -Smirnov)
- Pearson Chi-squared
- Shapiro Francis
They sound like fancy cocktails. Unless you're a stats buff, you can pick a couple that are familiar to you/ your discipline to run your data through (it literally takes 3 seconds- see the tutorial below).. and Voila!
STATS - this is the BIG GUNS package. More functions that I want to elucidate here. Just click the link.
This is where you will find such classic tests like:
- T-test
- Mann Whitney
- Tests of Variance
- ANOVA
- Tukey test
- distributions: gamma, normal, logistic, log normal....
- residuals
- principle components analysis
- GLM
- .....etc
GGPLOT - make pretty graphics worthy of the Economist! Here's a full on tutorial hosted by Harvard.
# Install GGPLOT package, then load it.
> install.packages("ggplot2")
> library("ggplot2")
Before you begin, save wing.csv and ABC Script 2 Tests for normality in a folder on your desktop.
A how to install packages web page.
A List of Lists
A list of commonly used R packages
A list of all shared R packages (alphabetical)
Google Sites
Report abuse