QERM 514 meets Tuesday and Thursday 10:30 to 11:50 in FSH 109. We will also be having weekly (optional) lab sessions on Fridays from 1:30-3:30 in FSH 209 starting on 4/4/03 to provide assistance with the computer assignments.
What's New?
Here is the deer jaw data from class on 5/29/03 related to performing non-linear regression. Also, here is a short, un-annotated script to perform the analysis of this data.
Labs
There won't be lab this week (5/30/03), but I think Prof. Conquest will hold office hours during the lab time.
Lab #8 will be held on 5/23/03 in Fish 209. Again, I'll be there to answer questions, etc. but there won't be a prepared script. Here is the script file, which I think is going to be presented in class next week.
Lab #7 will be held on 5/16/03 in Fish 209. I'll be there to answer questions, etc. but there won't be a prepared script. Here is the lizard data if you are interested in playing with it.
Lab #6 will be held on 5/9/03 in Fish 209. We will go over more about binomial glms (more on plotting, looking at residuals, etc.)and then do a quick introduction to poisson glms and functions such as update(), addterm() and dropterm() for 'quickly' analyzing these models. The files you need for this lab are: the lab6 script file and the leukemia data file.
Lab #5 will be held on 5/2/03 in Fish 209. We will go over some quick hints about managing contrasts and then have an introduction to binomial glms. The files you need for this lab are: the lab5 script file and the budworm data from V&R.
Lab #4 will be held on 4/25/03 in Fish 209. The purpose of this lab will be to go over how to run an ANCOVA, including setting the up and evaluating the contrasts, and interpreting the results. Also, we'll be covering a number of new plotting functions. The files you need for this lab are: the lab4 script file, the gen.data4 function, and the caterpillar data from class.
Also, here is the grazing data from class if you want to play with it.
Lab #3 will be held on 4/18/03 in Fish 209. The purpose of this lab will be to go over residual diagnostics and use of the box cox method to determine the appropriate transformation for the response variable. The files you will need are: the lab3 script file, the woodbeam data from class, and the gen.data3 function. Also, just in case, here is the alternate script for the boxcox procedure.
Lab #2 will be held on 4/11/03 in Fish 209. The purpose of this lab will be to provide an example of testing for lack of fit using ANOVA and as.factor(), to use tapply() to perform functions on groups of data, to learn some of the ways to tweak the plot() command, to identify points on a plot interactively, and to introduce the update() command to perform multiple linear regression. The files you will need are: the lab2 script file, the electric fish data, and the gen.data2 function.
Lab #1: We covered a quick intro to S-Plus and then went over generating random data, analyzing ordinary linear models using the lm() function, plotting simple data, creating and manipulating (transpose, inverse, multiply) matrixes, and analyzing gaussian/normal models using the glm() function. Look here for a summary of the commands we used and here for a script function to generate random data. These can both be opened in either directly in S-Plus or in any text editor.
Other notes and info
(4/8/03) A script file containing information about how to read the data files into an S-Plus object using read.table() (same information as was in the e-mail) is located here.
- If you haven't had much experience with S-Plus, here is an Introduction to S-Plus from Maureen Kennedy (last year's TA). I highly recommend that you at least browse through this before coming to lab.
Please e-mail me at: nhl@u.washington.edu if you have any S-Plus quesions!
Last Updated: 5/29/03