Hi, my name is Pascal Irz. I am an ecologist / datascientist at the French Biodiversity Office (OFB). In this position I mainly process fauna records for database quality enhancement, reporting, as well as to produce some ecological knowledge.
Reproducibility allows me to share developments with my colleagues in other regions or institutions as well as to get feedback and contributions. So, I try, as much as possible, to avoid « one shot » studies of local interest, and, instead, aim at the development of toolboxes with multiple uses.
I have a strong feeling that working according to the FAIR principles is interesting in research, but also in engineering, so I promote the use of R / RMarkdown + versioning by my non-researcher colleagues.
Cheers.
Pascal
Hi everyone,
My name is Jean-Christophe Loiseau. I am a Maître de Conférences in Applied Mathematics and Fluid Dynamics in the DynFluid laboratory which is part of Ecole Nationale Supérieure d’Arts et Métiers.
I received my PhD in 2014 from ENSAM. The main objective was the development of iterative solvers for large-scale eigenvalue problems with a particular emphasis on hydrodynamic stability analysis to better understand the early stages of transition to turbulence in fully three-dimensional flow configurations. Shortly after, I went to Politecnico di Bari to work on nonlinear optimization problems once again with an emphasis on better understanding the transition to turbulence. It was followed by a couple of years at the Royal Institute of Technology of Sweden (KTH) where I extended these approaches for multiphase flows.
Since 2017, I hold a Maître de Conférences back at Arts et Métiers. The core of my research is the development of reduced-order modeling techniques for real-time flow control. This includes both scaling up classical techniques from linear dynamical system theory as well the use of statistical learning techniques to identify nonlinear ordinary and partial differential equations directly from data. I am one of the core developers of
-
nekStab: an open-source fortran toolbox to perform bifurcation analysis using the computational fluid dynamics solver Nek5000.
-
PySINDy: a python library implementing several sparse regression techniques for the identification of ODE and PDE from data.
-
LightKrylov: a lightweight implementation of Krylov techniques using abstract types in Fortran.
I mostly work with Python, Fortran and Julia as programming languages. I am a strong believer of the Church of Emacs and use #org-mode quite a bit. I have also started to look at #Quarto recently. Like everybody here, I am very interested into pratical solutions to ensure computational reproducibility and tracability.
Léopold Carron, data scientist à L’Oréal, qui trouve dommage que ce réseaux soit très tourné universitaire, tout ce qui en sortira semble passionnant!
Sinon, pour me résumer : thèse et post doc en bioinformatique, et maintenant je suis plus proche de ce qu’on appellerais un ‹ computational toxicologist ›.
Hi all,
I am Romain Caneill, I did my PhD in Gothenburg (Sweden) and I am now in postdoc in Grenoble. I study physical oceanography, mainly using numerical models, but also with observations. I am a free software enthusiasts, and I think that reproducibility should be the number 1 rule of good science.
I have for now published 2 paper as 1st author, and both of them are (hopefully) 100% reproducible, with the help of snakemake and apptainer, which is something I am very proud of.
So I joined the recherche-reproductible forum and mainling list in the hope of meeting people, and stay informed on the state of the art in France.
I also wrote a simple python package that is simplifying the reading the outputs of the numerical ocean model NEMO: GitHub - rcaneill/xnemogcm: Interface to open NEMO global circulation model output dataset with xarray and create a xgcm grid.