Peihua Qiu's Software

Note: I received many requests for computer codes to implement the procedures
suggested in the papers listed below. For readers' convenience, some of these
codes have been reasonably documented and included here for free download.

• Statistical process control (SPC) for multivariate data that is not
  necessarily Gaussian distributed. A function is provided to categorize
  components of multivariate response vectors. Tools for setting up a
  CUSUM procedure for the transformed data are included. The CUSUM scheme
  can also be applied to the case when some (or all) of the multivariate
  response components are binary-categorical.
  • Qiu, P., "Distribution-free multivariate process control based
    on log-linear modeling," IIE Transactions (a flagship journal of
    the Institute of Industrial Engineers), 40(7), 2008, 664--677.
    ( ps file, pdf file)
  • R package "mnspc".
  • This paper was featured by the article published in the
    Industrial Engineer magazine, 40(7), 2008, 53--54.
    

• In survival analysis applications, we often need to compare two hazard
  rate functions to investigate the effectiveness of a new treatment,
  compared to a standard treatment (or a control). Quite often, the two
  hazard rate functions cross each other, due to the facts that certain
  treatments have significant effects only after/before certain time periods.
  For instance, radiation and chemotherapy can usually improve patients'
  prospects for short-term survival; but they have little or no long-term
  medical benefits. Surgery, on the other hand, has benefits in the long run;
  it may increase the risk in the early stage after it is applied. To compare
  two possibly crossing hazard rate functions, we suggested a two-stage
  procedure in the following paper.
  • Qiu, P., and Sheng, J., "A Two-Stage Procedure For Comparing
    Hazard Rate Functions,"JRSS-B, 70(1), 2008, 191-208.
    ( ps file, pdf file)
  • R package "TSHRC" can be downloaded from the R web page
    under the link "Packages" on the left-hand side of the web page.
    

• We derived formulas for sample size calculation for a pair-matched
  case-control study to test for interaction between a specific exposure
  and a second risk factor. The exposure variable considered is binary
  but the second risk factor could be either binary or continuous. The
  cases in which the second risk factor is OR is not a matching variable
  are discussed separately. This research is motivated by a study
  comparing the prevalence of GP-IIIa polymorphism (the exposure) in
  individuals with and without myocardial infarction (case-control).
  • Qiu, P., Moeschberger, M., Cooke, G., and Goldschmidt-Clermont, P.,
    "Sample size to test for interaction between a specific exposure and a
    second risk factor in a pair-matched case-control study", Statistics in
    Medicine 19, 2000, 923-935.
    ( ps file, pdf file)
  • Two Splus programs: figure3.s, figure4.s.
    

• We developed a statistical modeling procedure for sleep-wake patterns
  in response to light-dark cycles of varying lengths, which would take
  into account both the rhythmic circadian component and stimulus-induced
  shifts in behavior. Such a model could also be used to describe other physiological
  parameters which may include both rhythmic and acute components, such as
  temperature and hormone secretion.
  • Qiu, P., Chappell, R., Obermeyer, W., and Benca, R., "Modelling daily
    and subdaily cycles in rat sleep data", Biometrics 55, 1999, 930-935.
    ( ps file, pdf file)
  • Two datasets: baseline.dat, test.dat.
  • Splus program.
    

Peihua Qiu's Home Page