Embedding R in Windows applications, and executing R remotely

More documents

Recommendations

Info

On Multiple Comparisons in R by Frank Bretz 1 , Torsten Hothorn 2 and Peter Westfall 3 1 Universität Hannover, Lehrgebiet Bioinformatik, Hannover 2 Friedrich-Alexander-Universität Erlangen-Nürnberg, Institut für Medizininformatik,Biometrie und Epidemiologie 3 Texas Tech University, Dept. of Information Systems and Quantitative Sciences, Lubbock, TX The multcomp package for the R statistical environment allows for multiple comparisons of parameters whose estimates are generally correlated, including comparisons of k groups in general linear models. The package has many common multiple comparison procedures “hard-coded”, including Dunnett, Tukey, sequential pairwise contrasts, comparisons with the average, changepoint analysis, Williams’, Marcus’, McDermott’s, and tetrad contrasts. In addition, a free input interface for the contrast matrix allows for more general comparisons. The comparisons itself are not restricted to balanced or simple designs. Instead, the programs are designed to suit general multiple comparisons, thus allowing for covariates, nested effects, correlated means, likelihood-based estimates, and missing values. For the homoscedastic normal linear models, the program accounts for the correlations between test statistics by using the exact multivariate t-distribution. The resulting procedures are therefore more powerful than the Bonferroni and Holm methods; adjusted p-values for these methods are reported for reference. For more general models, the program accounts for correlations using the asymptotic multivariate normal distribution; examples include multiple comparisons based on rank transformations, logistic regression, GEEs, and proportional hazards models. In the asymptotic case, the user must supply the estimates, the asymptotic covariance matrix, and the contrast matrix. Basically, the package provides two functions. The first one computes confidence intervals for the common single-step procedures (simint). This approach is uniformly improved by the second function (simtest), which utilizes logical constraints and is closely related to closed testing. However, no confidence intervals are available for the simtest function. In this talk we give an introduction to the multcomp package. We first provide a brief theoretical background on multiple comparisons and the multiple contrast representation. We then illustrate the use of the package by going through several examples.
Using R for Statistical Seismology Ray Brownrigg Statistical Seismology is a relatively new field which applies statistical methodology to earthquake data in an attempt to raise new questions about earthquake mechanisms, to characterise aftershock sequences and to make some progress towards earthquake prediction. A suite of R packages, known as SSLib, is available for use with Statistical Seismology in general, but also with some applications outside this field. This presentation will introduce the packages, describe the way certain special features of R have been crafted to a framework for researchers in the field and demonstrate some of the exploratory data analysis functions available within the packages.
Page 1 and 2: Embedding R in Windows applications
Page 3 and 4: Using R in a Distributed Computer L
Page 5: hoa - A package bundle for higher o
Page 9 and 10: Ô5Õ‚Ö/× ØgÙL× Ú#Õ‚Û$
Page 11 and 12: Calculating the autocovariances of
Page 13 and 14: Using R for predicting air traffic
Page 15 and 16: Deployment of validated statistical
Page 17 and 18: RMAGEML: integrating MAGE-ML format
Page 19 and 20: Given the large, and growing, exten
Page 21 and 22: It is our intention to extend the m
Page 23 and 24: Programming with financial data: Co
Page 25 and 26: 2.1 The foundation provided by Knop
Page 27 and 28: • permits the choice for each nod
Page 29 and 30: Computational cost and time threads
Page 31 and 32: Getting Started With the R Commande
Page 33 and 34: Bayesian Methods in Geostatistics:
Page 35 and 36: The Extreme Value Toolkit (extRemes
Page 37 and 38: BayesMix: An R package for Bayesian
Page 39 and 40: The Statistical Lab - Teaching and
Page 41 and 42: PRSG Nirvana — A parallel tool to
Page 43 and 44: Microeconomic Analysis with R Arne
Page 45 and 46: Multi-state Markov modelling with R
Page 47 and 48: metrics: Towards a package for doin
Page 49 and 50: Performance of Spinal Supporting Mu
Page 51 and 52: Distributions of finite and infinit
Page 53 and 54: RUnit - A Software Development And
Page 55 and 56: SciViews package The SciViews packa
Page 57 and 58:
An Adventure in randomForest - A Se
Page 59 and 60:
Simulating Society in R András Lö
Page 61 and 62:
MCMCpack: An Evolving R Package for
Page 63 and 64:
Orientlib: Using S Version 4 Object
Page 65 and 66:
Taking into account spatial depende
Page 67 and 68:
which correctly records in the char
Page 69 and 70:
useR! - at an investment bank? Bern
Page 71 and 72:
userR!2004 Conference Abstract R at
Page 73 and 74:
Doing Customer Intelligence with R
Page 75 and 76:
Option pricing using R John Randal
Page 77 and 78:
Bioassay analysis using R Christian
Page 79 and 80:
klaR: A Package Including Various C
Page 81 and 82:
An empirical Bayes method for gene
Page 83 and 84:
Multiple Imputation and Model Selec
Page 85 and 86:
a. b. c. d. Figure 2: Influence by
Page 87 and 88:
Rho: An Integrated System for Rba
Page 89 and 90:
JGR: a unified interface to R Marti
Page 91 and 92:
RNGstreams — An R package for mul
Page 93 and 94:
Rough Set based Rule Induction Pack
Page 95 and 96:
Using R in other Applications Vario
Page 97 and 98:
Using R and SPSS simultaneously in
Page 99 and 100:
[5] J. Herrero, R. Díaz-Uriarte, a
Page 101 and 102:
¢¡¤£¦¥¨§©££¨£¤¡ !¡"
Page 103 and 104:
Propensity score analysis 2 We will
Page 105 and 106:
USE R FOR PEPTIDE MASS FINGER PRINT
Page 107 and 108:
Rmetrics www.rmetrics.org An Enviro
show all

Embedding R in Windows applications, and executing R remotely

Create successful ePaper yourself

Delete template?

Save as template?