flowchart LR A["Methodological research"] --> B["R packages"] B --> C["Applications and diagnostics"] C --> D["Reproducible research repositories"] D --> E["Teaching resources"] E --> F["Open dissemination"] A --> A1["GLBFP"] A --> A2["Validation-SSF"] B --> B1["ggcircular"] B --> B2["CircularRegression"] B --> B3["gmov"] C --> C1["contextR"] E --> E1["donnees-bleues"] E --> E2["tutorizeR"]
Research Lab
Open research ecosystem in computational statistics, scientific R software, reproducible data science, and statistical pedagogy.
This page presents the research, scientific software, and open pedagogy ecosystem developed by Aurélien Nicosia at Université Laval. It connects GitHub repositories, R packages, teaching resources, and reproducible research materials in computational statistics.
Mission
To develop an open platform for computational statistics that connects methodological research, scientific software, reproducible data science, and statistical pedagogy. The aim is to produce interpretable methods, usable R tools, and open resources that support transmission, verification, and reuse.
Research areas
Densities and anomalies
Density estimation and anomaly detection
Density estimation methods based on the General Linear Blend Frequency Polygon approach, with attention to stability, geometric interpretation, and diagnostics for atypical observations.
Directional data
Circular and directional statistics
Models, visualizations, and tools for angles, orientations, directions, times of day, and circular responses.
Animal movement
Movement modeling, HMMs, SSFs, and generative validation
Hidden Markov models, step-selection functions, iSSF models, and generative validation of step-selection functions to evaluate simulated movement behavior.
Interpretation
AI-assisted statistical interpretation
Tools and workflows for contextualized statistical explanations, with methodological safeguards and attention to interpretation quality.
Open pedagogy
Open statistical pedagogy and reproducible data science
Quarto resources, pedagogical datasets, interactive tutorials, and reproducible environments for teaching statistics, R, and data science.
Ecosystem map
Featured projects
Densities and anomalies
GLBFP
R package for density estimation using the General Linear Blend Frequency Polygon approach. It is the methodological core of the density and atypical-observation axis.
Directional data
ggcircular
ggplot2 extension for visualizing circular, axial, and directional data. It supports graphical communication for methods involving angular data.
Directional data
CircularRegression
R package for fitting circular-linear regression models, especially for directional and trajectory-related data.
Animal movement
Validation-SSF
Reproducible research repository for generative validation of step-selection functions, including simulations, empirical analyses, and sensitivity analyses.
Animal movement
gmov
Generative validation and diagnostic tools for SSF and iSSF movement models. The project connects methodological work with practical movement-model diagnostics.
Data and pedagogy
donnees-bleues
Open educational platform of Québec datasets for statistics and data science. It provides realistic data for teaching and reproducible exercises.
Open pedagogy
tutorizeR
R package for converting R Markdown and Quarto documents into interactive learnr tutorials with exercises, solutions, and optional quizzes.
Emerging project
contextR
Emerging project around contextualized statistical explanations in R, with S3 structures, safeguards, and links to AI-assisted statistical interpretation.
Scientific software
The R packages form a central layer of the ecosystem. They turn methodological ideas into tools that can be tested, documented, and reused. GLBFP, ggcircular, CircularRegression, gmov, tutorizeR, and contextR cover density estimation, circular visualization, circular-response modeling, generative movement validation, interactive tutorial production, and contextualized statistical interpretation.
Data and pedagogy
The teaching resources support open, realistic, and reproducible training in statistics and data science.
Datasets
donnees-bleues gathers Québec datasets for courses, exercises, and projects.
Courses and modules
The course and resource sites structure content in statistics, R, and data science.
Interactive tutorials
tutorizeR supports the conversion of reproducible documents into interactive teaching resources.
Reproducible research
Reproducible research repositories document code, analyses, figures, simulations, and materials associated with scientific work. This layer makes results easier to verify and reuse by collaborators, students, article readers, and scientific software users.
Collaborations and contact
Collaborations are welcome on statistical methods, R software, movement ecology applications, directional data, statistical pedagogy, and reproducible data science.