PubH 7460 Biostatistical Computing II - Fall 2013


School of Public Health


Division of Biostatistics


 It is a good thing for an uneducated man to read books of 
quotations.
 - Winston Churchill

A day without sunshine is like, you know, night. - Steve Martin

I believe that a scientist looking at nonscientific problems is just as dumb as the next guy. - Richard Feynman

If you can't explain it to a six year old, you don't understand it yourself. - Albert Einstein


The intent in this course is to gain some skill in statistical computing using SAS, R, Splus, and possibly FORTRAN or C. The recurring theme is solving problems by turning an algorithm into a program that provides relevant answers. Some background in probability and statistics, including an introduction to likelihood, is needed.

Topics covered will include use of pseudo-random number generators, distribution functions (pdfs and cdfs), matrix manipulations with applications to regression and estimation of variance, simulation studies, minimization of functions using various algorithms, nonlinear regression, manipulation and combination of datasets, techniques of macro programming, and methods of integration using quadrature.

The course is intended primarily for MS and MPH students in Biostatistics and Statistics. At least one previous semester of courses in biostatistics and statistics and concurrent or previous enrollment in a course involving linear regression are strongly recommended.

The main content of the course is a series of computing projects (see Syllabus). Datasets based on previous studies will be available for providing examples of computational methods.

This course is not intended as a continuation of SPH 6420. In general the level of statistical knowledge and mathematical skill is somewhat advanced over that required in SPH 6420 or the SPH 6450-6451 sequence.


  • Categorical Data Lecture 15

  • Categorical Data Lecture 16

  • kisumu.2011

  • SAS analysis of crossover data ...

  • Excerpt from Cochran & Cox ...

  • Kutner et al. Table 28.11: Apple Sales
  • Three-Period Crossover Example ...
  • Incomplete Block Design Example ...
  • The Two Kinds of Orthogonality ...
  • Orthogonality and Un-Correlation ...
  • Split-Plot Example: Snedecor & Cochran ...
  • SAS Program for the Snedecor & Cochran example ...

  • EXCEL FILES FOR AFRICA 2010:

  • Biostat Intro 1 Excel File
  • Biostat Intro 2 Excel File
  • Biostat Intro 3 Excel File
  • Biostat Intro 4 Excel File

    Notes for this course:

  • notes.001 : Random number generators
  • notes.001.1 : Useful Functions
  • notes.002 : Randomization schedules
  • notes.003 : Pseudo-random numbers
  • notes.004 : Sample size, power, simulations
  • notes.004.1 : Use of a SAS macro to call a procedure repeatedly in simulations ...
  • Excerpt from Lachin: Sample Size for Survival Analysis
  • notes.005 : Binomial outcomes
  • notes.006 : Some program examples
  • notes.007 : More on simulations
  • Numeric Computation of PDFs / Derivatives.
  • notes.008 : Computation of summary stats
  • --- lhs.listing : datafile for Prob 10.2, notes.008
  • mediansum.examp : Computation of Median, Use of SYMPUT.
  • notes.009 : More on summary stats
  • notes.010 : More on summary stats
  • --- lhs.data : datafile for Prob 11, notes.010
  • notes.011 : Mysteries of the SAS data step
  • notes.012 : Simulation from additional distribs
  • regression.problem: Simulation to Answer a Regression Question
  • asthma.out: Data Set for Homework Assignment on Permutation Test
  • notes.013 : Linear transformations, matrices 1
  • Linear transformation of unit square.
  • notes.014 : Linear transformations, matrices 2
  • notes.015 : Linear transformations, matrices 3
  • notes.016 : SAS proc iml
  • notes.016a : SAS proc iml basics ...
  • notes.016b : Bootstrap Using PROC IML
  • notes.017 : Proc iml: solving nonlinear eqns
  • notes.018 : Proc iml: finding maxima, minima
  • notes.019 : Proc iml: maximum likelihood estn
  • notes.020 : Proc iml: max like estn, contin.
  • PROC NLP applied to regression with truncated data
  • notes.021 : Proc iml: simulating multivar normals
  • notes.021a : Using Eigenvectors to Define Uncorrelated Variables
  • notes.022 : SAS macros: macro variables
  • notes.023 : SAS macros: basics plus ...
  • notes.023a : SAS macros: Some generalities & examples
  • notes.023b : SAS macros: Simulations using SAS procedures
  • notes.024 : SAS macros: SAS/GRAPH ...
  • notes.025 : SAS macros: More SAS/GRAPH ...
  • notes.026 : SAS PROC NLIN: Introduction.
  • notes.emalgorithm : EM Algorithm.
  • Zero-Inflated Poisson Data: 3 approaches
  • Random Permutations and Permutation Tests.
  • notes.027 : SAS PROC NLIN and amoeba ...
  • notes.028 : Variance Estimation, I
  • notes.029 : Variance Estimation, II : Delta Method
  • notes.030 : Numerical Integration
  • notes.031 : Numerical Integration, Contin.: Romberg Integration
  • notes.032 : Bayes Computations and Numerical Integration


  • Biostatistics Home Page

    Web address of this page: http://www.biostat.umn.edu/~john-c/ph7460.f2006.html

    Most recent update: December 2, 2013.


  • Cape Buffalo, Lake Nakuru, Kenya, August 2010

    For additional photos from Africa and others, see: John C. Flickr Website

    Web counter provided by WebCounter(TM), http://www.digits.com