Share on Facebook Tweet on Twitter Share on LinkedIn Share by email
Distribution Modeller

Since its inception, the Computational Science group has undertaken research and development into new modelling platforms for computational science. The CEESDM project detailed here evolved from the Computational Science Studio project (mentioned in this article) and then evolved into the Altai project. The Altai project then evolved into the Modelling Environment project, the 4th generation of modelling environment being developed by the group (coming soon...).

Platform and Computational Experiment for Smith et al. (2015) "Inferred support for disturbance-recovery hypothesis of North Atlantic phytoplankton blooms".

Recently Matthew Smith, Vassily Lyutsarev, Derek Tittensor and Eugene Murphy published the first complete peer-reviewed piece of scientific research using Altai. That paper includes a link to this page for people wanting to use the computational platform used to generate the results for that study and the complete computational experiment (the computational steps that led to the results)

  • To install the computational platform used for that study go to the Altai page (note the prerequisites on that page)
  • The computational experiment, which can be accessed using Altai, is available on GitHub from this page.

Archive of information about Distribution Modeller

Distribution Modeller (temporary name only!) was CEES' end-to-end browser tool that let the researcher rapidly import data, supplement that data with environmental info from FetchClimate, specify an arbitrary model by point and click or in code, parameterize the model against the data using Filzbach, make and visualize predictions with a full propagation of parameter uncertainty – then package and share everything, in a way that is inspectable, repeatable, and modifiable. Most of these features were carried forward into the subsequent modelling environment development projects

Distribution modeller was a CEES project in association with Microsoft Research Connections.