The R package miCoPTCM, written by Aurelie Bertrand for the analysis of cure models in survival analysis, implements the programs used in (a) Bertrand A., Legrand C., Léonard D. and Van Keilegom I. (2017). Robustness of estimation methods in a survival cure model with mismeasured covariates Computational Statistics and Data Analysis 2017, in press; and (b)
Bertrand, A., Legrand, C., Carroll, R. J., de Meester, C. and Van Keilegom, I. (2017). Inference in a survival cure model with mismeasured covariates using a SIMEX approach. Biometrika, to appear.
R code for implementing the misclassification analysis in the paper Cook, S. J., Blas, B., Carroll, R. J. and Sinha, S. (2017). Two wrongs make a right: addressing underreporting in binary data from multiple sources. Political Analysis, to appear
Matlab and R code for implementing the dietary patterns methodology in the paper
Ma, S., Ma, Y. Wang, Y., Kravitz, E. S. and Carroll, R. J. (2017). A semiparametric single-index risk score across populations. Journal of the American Statistical Association, to appear.
BayesME: an R package for nonparametric density deconvolution and nonparametric regression (not yet implemented) allowing for heteroscadastic measurement error and heteroscedastic regression error .Based on Sarkar, A., Mallick, B. K., Staudenmayer, J., Pati, D. and Carroll, R. J. (2014). Bayesian semiparametric density deconvolution in the presence of conditionally heteroscedastic measurement errors. Journal of Computational and Graphical Statistics, 25, 1101-1125 and Sarkar, A., Mallick, B. K. and Carroll, R. J. (2014). Bayesian semiparametric regression in the presence of conditionally heteroscedastic measurement and regression errors. Biometrics, 70, 823-834. Packages mvtnorm and msm are required.
Programs for the method of secondary linear regression analysis of case-control studies. The paper, Semiparametric Estimation in the Secondary Analysis of Case-Control Studies by Yanyuan Ma and Raymond J. Carroll, will appear in the Journal of the
Royal Statistical Society, Series B, in 2015. There is a README file and an example data set to use as a test.
CGEN, an R packages based on my work with Nilanjan Chatterjee and Yi-Hau Chen for analyzing genetic data on case-control samples, with particular emphasis on novel methods for detecting Gene-Gene and Gene-Environment interactions.
SAS Macro and Demonstrations For Estimating Usual Intake Distributions of Food and Associated Individual-Level Predictors for Diet-Disease Relationships
This site is based on our work involving the estimation of the usual intake distributyion for nutrients and episodically consumed foods. In addition, we estimate individual-level usual intake for use in regression calibration analysis of diet-disease relationships.
SAS Macro for Haplotype Analysis
The SAS macro HapReg implements the haplotype-based genetic association
analysis for case-control studies, using a flexible model for gene-environment
association allowing haplotypes to be potentially related with environmental
exposures. The novel methodology is proposed by Chen, Chatterjee, and Carroll
(2007, Retrospective Analysis of Haplotype-Based Case-Control Studies Under a
Flexible Model for Gene-Environment Association, under revision).
Wavelet-Based Functional Mixed Model Methodology, Windows
These are programs based upon the work of Jeffrey Morris and Raymond Carroll, 2006, Journal of the Royal Statistical Society, Series B, 68, 179-199
MatLab and R Programs for Mixtures of Berkson and Classical Measurement Errors in the Nevada Test Site Study
These are programs based upon the work of Yehua Li, Annamaria Guolo, F. Owen Hoffman and Raymond J. Carroll. There are Bayesian and Monte-Carlo EM programs for the analysis of these important radiation data.
Programs called in the Measurement Error Short Course at ENAR, 2008
These are Stata, R2WinBUGS and some SAS Code for our measurement error short course.
The following programs are in zipped directories. The various utility directories need to be unzipped since some of the functions call them.
Raymond Carroll’s Utility Directory
Directory is called “carroll_utilities_directory”
David Ruppert’s Utility Functions
Directory is called “utility”.
SIMEX and Bayes Programs for Gaussian and Logistic Regression in the Instrumental Variables Problem
Uses regression splines. The set of programs is self-contained, i.e., call no other of my directories. Based on a paper Nonlinear and Nonparametric Regression and Instrumental Variables by Raymond J. Carroll, David Ruppert, Ciprian M. Crainiceanu, Tor D. Tosteson and Margaret R. Karagas.
SIMEX and Bayes Programs for Gaussian Regression (linear link) with classical measurement error
The README files described which directories to call, namely the “carroll_modified_spline_functions” and “carroll_utilities_directory”. Based on the paper Bayesian smoothing and regression splines for measurement error problems by S. Berry, R. J. Carroll and D. Ruppert, JASA, 97, 160–169.
Spline and Kernel Methods for Longitudinal Data (Panel structure, linear link)
Directory is called “panel_data_directory”. Computes regression spline estimators using GLS and working independence, and two–stage kernel estimators. Based on the paper Linton, O. B., Mammen, E., Lin, X. & Carroll, R. J. (2003). Accounting for correlation in marginal longitudinal nonparametric regression. Second Seattle Symposium on Biostatistics, editor D. Lin. You must call the directories “srs”, “lpoly”, “ebbs”, “utility”.
Carroll Regression Spline Functions that have nothing to do with measurement error
Directory is called “carroll_modified_spline_functions”
Programs for local polynominal regression with bandwidth estimated from EBBS (A David Ruppert method)
Directory is called “ebbs”.
Programs for spline regression that combine some of mine with some of David Ruppert’s
Directory is called “srs”.
Programs for local polynomial regression that combine some of mine with some of David Ruppert’s
Directory is called “lpoly”.
Programs for Bayesian fitting of adaptive P-splines in regression
Directory is called “adaptive_psplinesv06”. You must call the directories “srs”, “lpoly”, “ebbs”, “utility”.
Programs for nonparametric regression in clustered data binning
Programs are self–contained, i.e., call no other directories. Based upon the paper A Simple, General Approach To Inference In Mixed Parametric And Semiparametric Models by R. J. Carroll, P. Hall, T. V. Apanasovich and X. Lin.
Programs for analysis of matched case-control data with missing observations
Based on the paper Semiparametric Bayesian Analysis of Matched Case-Control Studies with Missing Exposure by S. Sinha, B. Mukherjee, M. Ghosh, B. K. Mallick and R. J. Carroll
Taylex and Other Deconvolution Programs
This implements the Taylex method of Carroll and Hall, JRSSB, 2004, volume 66, pages 31-46. Calls utility, loply, ebbs and carroll_utilities.
John Staudenmeyer’s programs for SIMEX in nonparametric regression with efficient bandwidth estimation
Based on the paper Staudenmeyer, J. and Ruppert, D. (2003). Local polynomial regression and SIMEX.
The R programs are basically translations of the MATLAB Programs. They are all in one directory.
R programs, zip file
Directory is called “R_programs”