Other EDA Examples
Use these examples as a starting point for your own explorations
of some of the methods discussed in the lectures.
Many are quite short. Feel free to experiment, and call for help
if you'd like more explanation.
You can use
%webhelp(macroname);
to access the Web documentation for most (but not all) of the
macro programs.
Assessing bivariate relations
- Data ellipse for a bivariate relation.
[File: contiris2.sas]
- Baseball data: Smoothing Salary vs. Years
[File: baseyear.sas]
- Brain weight and body weight: Ratio of Slopes table
[File: brains.sas]
Assessing multivariate relations
- Examining multivariate normality with the MULTNORM macro.
[File: multnorm1.sas]
- Outlier demonstration: 2 & 3 variables.
[File: outldem1.sas]
Illustrates the problem of multivariate outliers in simple 2- and 3-variable
situations, where univariate or even bivariate plots may not show anything unusual.
- Outlier detection: Simple ChiSquare Probability plot.
[File: outldem2.sas]
A continuation of OUTLDEM1, this example shows the basic idea behind the
outlier plot.
Missing data
- Some simple (but poor) ways to handle missing data -- unconditional means,
conditional means.
[File: automiss.sas]
- Missing data imputation with PROC FASTCLUS -- cluster means.
[File: impute.sas]
- Multiple imputation using the MISS and MISSCOMB macros.
[File: basemiss.sas]
- Multiple imputation using PROC MI and PROC MIANALYZE (monotone missing)
[File: basemi.sas]
- Multiple imputation using PROC MI and PROC MIANALYZE. (non-monotone missing)
[File: fitmi.sas]