Software

I am an active member of the R software community and was elected to the R Foundation for Statistical Computing.

R Package Development

  • FactoMineR: Exploratory data analysis (PCA, MCA, multi-table analysis, etc.).
  • missMDA: Imputation (matrix completion) for continuous/categorical data and PCA with missing values.
  • denoiseR: Low-rank matrix estimation using regularized SVD and bootstrap.

My students have also developed packages linked to our work:

  • misaem: Logistic regression with missing values.
  •  mimi: Generalized low-rank models for mixed/incomplete data.
  • lori: Contingency tables with missing values and covariates.
  • AdaptiveConformal: Adaptive conformal prediction for time series.

For causal inference with missing values, see grf and pipelines comparing estimators (IPW, AIPW, etc.).

Software Projects

  • R-mis-static: [paper] A static website providing resources on missing data (lectures, workflows, tutorials). Contributions welcome!
  • Causal Inference Taskview: Organizes R packages related to causal inference.

ICUBAM Development

I contributed to ICUBAM, an open-source tool for real-time visualization of ICU bed availability. Launched during the COVID-19 crisis, it aids ICU teams in managing patient flows, anticipating bed needs, and organizing transfers. Deployed in 130 ICUs across 40 départements, ICUBAM supports over 2,000 beds. Slides, Models, Paper, GitHub.

Community Engagement

  • Associate Editor, Journal of Statistical Software (2011–2017).
  • Founding member of RForwards, promoting diversity in the R community.
  • Member of the R Foundation Conference Committee; contributed to implementing the Code of Conduct.
  • Co-founder of the « French R Board » supporting Les Rencontres R.

Help R by supporting with donation or through the R consortium.