Software

This page gathers some R packages, set of R functions or shiny applications available on local shiny server. Most of them are commented and should work without any problems. Should you encounter some troubles, please feel free to contact me.

Shiny applications

  • benford_presidentielles: to illustrate Benford distribution on French presidential voting data (B Amistadi, C Champ, M Darrouzes, A Grancher and A Sabhou (L3 MI students))

  • E2019-2020: analyzes data from (the best ultra-trail ever) Echappée Belle (L Favre, D Ferrero, K MacKenna and D Tranchant (M1 Statitics and Data Science))

  • descriptionApp: summary/graphs of a few datasets we use in STA331 (T Silvestre)

  • dispersionApp: to understand/play with the concepts of location, dispersion, asymmetry, boxplots (T Silvestre)

  • distributionApp: to understand/visualize main probability distribution functions, cdf, quantile functions (T Silvestre)

  • CIapp: to understand confidence intervals for the theoretical mean from a Gaussian sample (T Silvestre)

  • drHouseApp: to understand/visualize all concepts related to hypothesis testing (first type error, power functions,...) (T Silvestre)


Fractional Brownian motion and related stochastic processes

  • simGCP.R - Set of R functions dealing with the simulation of a circularly-symmetric fractiolnal Brownian motion, a modulated fractional Brownian motion and the estimation of the Hurst exponent of a circularly-symmetric fractional Brownian motion. Have a look at the Rmarkdown: demoCFGN.Rmd and its pdf version: demoCFGN.pdf. These functions are related to these two papers.

    1. J.-F. Coeurjolly and E. Porcu. Fast and exact simulation of complex-valued stationary Gaussian processes through embedding circulant matrix, to appear in Journal of Computational and Graphical Statistics, 2017. [arxiv]

    2. J.-F. Coeurjolly and E. Porcu. Properties and Hurst exponent estimation of the circularly-symmetric fractional Brownian motion, Statistics and Probability Letters, 128:21–27, 2017.

  • SimEstFBM.R Set of R functions dealing with the simulation of a fractional Brownian motion and the estimation of its parameters, related to the papers

    1. Coeurjolly Jean-François. Estimating the parameters of a fractional Brownian motion by discrete variations of its sample paths. Statistical Inference for Stochastic Processes, 4 (2), 199–227, 2001. [pdf]

    2. Coeurjolly Jean-François. Simulation and identification of the fractional Brownian motion: a bibliographical and comparative study. Journal of Statistical Software, 5 (7), 1–53, 2000.[pdf]

  • dvfBm (with S. Achard) Discrete variations for fractional Brownian motion. R package devoted to the robust estimation of the Hurst exponent. Additive outliers and Gaussian white noise are considered. This package is mainly related to

    1. Coeurjolly Jean-François. Hurst exponent estimation of locally self-similar Gaussian processes using sample quantiles. Annals of Statistics, 36 (3), 1404–1434, 2008. [pdf]

    2. Achard S. and Coeurjolly J.-F. Discrete variations of the fractional Brownian motion in the presence of outliers and an additive noise. Statistics Surveys, 4, 117–147, 2010.[pdf]

    3. J.-F. Coeurjolly and H. Kortas. Expectiles for subordinated Gaussian processes with applications.. Electronic Journal of Statistics 6, 304–323, 2012.[pdf]

  • Multivariate fractional Brownian motion: mfBm.R Set of R functions dealing with the simulation and the identification of the multivariate fractional Brownian motion. These functions are related to the papers

    1. P.O. Amblard, J.-F. Coeurjolly, F. Lavancier, and A. Philippe. Basic properties of the multivariate fractional Brownian motion. To appear in Bulletin de la Société Mathématique de France, Séminaires et Congrés, 28, 65–87 2012. [arxiv] [pdf].

    2. P.O. Amblard and J.-F. Coeurjolly. Identification of the multivariate fractional Brownian motion. IEEE Transactions on Signal Processing 59(11), 5152–5168, 2011. [pdf].

    3. J.-F. Coeurjolly,P.O. Amblard and S. Achard. Wavelet analysis of the multivariate fractional Brownian motion. To appear in ESAIM Probability and Statistics, 2012.[arxiv]

Statistics for spatial point processes

Implementation within the spatstat R package (A. Baddeley, E. Rubak and R. Turner)

  • confidence intervals for the mple (with E. Rubak) Within the 1-28 version, we implemented a generic function which aims at computing a very fast and computationally cheap estimate of the asymptotic covariance matrix of a very general stationary marked Gibbs point process. This work is mainly based on two papers

    1. Billiot J.-M., Coeurjolly J.-F. and Drouilhet R. Maximum pseudolikelihood estimator for exponential family models of marked Gibbs point processes. Electronic Journal of Statistics, 2, 243–264, 2008. [pdf]

    2. J.-F. Coeurjolly and E. Rubak. Fast estimation of covariances for spatial Gibbs point processes. Submitted, 2012.[Research report CSGB]

  • logistic regression estimate (with A. Baddeley, E. Rubak and R. Waagepetersen) Withing the 1-28 version, we implemented a new method, alternative to the pseudo-likelihood which is based on the logistic regression. This code implemented in the general ppm function of spatstat has many advantages compared to the pseudo-likelihood. The main one is that it is very fast, it does not require any disretization of integrals and like the pseudo-likelihood it can be implemented using the general glm (generalized linear model) function since the score of the estimating equation we propose look like the score of a logistic regression. Details can be found here

    1. A. Baddeley, J.-F. Coeurjolly, E. Rubak and R. Waagepetersen. Logistic regression likelihood for spatial point processes..

Miscellanea

    • asympTest (with R. Drouilhet, P. Lafaye de Micheaux and J.-F. Robineau) R package for performing parametric statistical tests and confidence intervals based on the central limit theorem. This package is related to the publication

      1. Coeurjolly J.-F., Drouilhet R., Lafaye De Micheaux P. and Robineau J.-F. asympTest: A simple R package for classical parametric statistical tests and confidence intervals in large samples. The R Journal, 1 (2), 26–30, 2009.[pdf]

    • Geodesic normal distribution on the circle: circGeod.R Set of R functions devoted to a new circular distribution: the geodesic normal distribution on the circle which can be viewed, roughly speaking, as the standard normal distribution where the euclidean distance is replaced by the geodesic one. These functions treat the problem of generating such a distribution and the problemn of estimating the parameters (and standard errors) using the maximum likelihood method. Details on this distribution can be found in this paper

      1. J.-F. Coeurjolly, N. Le Bihan. Geodesic Normal distribution on the circle. Metrika, 2012. [pdf]

    • Attributable mortality function using adjusted multi-state models: mortality.R (with B. Liquet) Set of R functions accompanying the paper cited below. These functions are concerned with the estimation of attributable functions using adjusted multi-state models. They have been developed during M. Nguile Makao's PhD who aimed at modelling the nosocomial pneumonia in intensive care unit of Grenoble. Details and dataset can be downloaded on journal's website.

      1. J.-F. Coeurjolly, M. Nguile-Makao and J.-F. Timsit, B. Liquet. Attributable risk estimation for adjusted disability multi-state models: application to nosocomial infections. Biometrical Journal, 54(5), 600–616, 2012. [pdf]