This site is deprecated. Please use the new site at https://pandora-isomemo.github.io/docs/ instead.

Bringing BIG Data & advanced computational modelling together

Pandora & IsoMemo Computational Lab

The Pandora & IsoMemo Computational Lab provides access to a diverse range of computational tools, including Artificial Intelligence (AI), for data querying, manipulation, and modeling. The lab encompasses various types of models across numerous R and Python applications. These models, either already available or under development, are designed for studying complex systems such as network analysis, Bayesian analysis of differential equations, Agent-Based modeling, and Bayesian power law analysis. Additionally, the lab offers tools for time series analysis, such as Bayesian counterfactual analysis, Bayesian Point Break analysis, and AI-based tipping point detection. Other features include tools for spatiotemporal modeling and different types of causal analysis, such as Bayesian networks, Bayesian regression, and Bayesian model selection algorithms.

The currently available Pandora & IsoMemo software can be accessed on the Pandora & IsoMemo software platform on GitHub. Each software repository provides detailed instructions on installation and usage. All software features a user-friendly interface developed using Shiny technology. Many of the applications offer innovative self-developed modeling. In some instances, user-friendly interfaces along with novel modeling features were developed in collaboration with the developers of modeling software packages. In other cases, user-friendly interfaces were designed for software packages available under open licenses.

Applications

Data Search

Allows users to access, query, and map data available from the Pandora platform and from its networks (e.g., IsoMemo)

Data Search and Spatiotemporal Modelling

In addition to allowing users to query Pandora data includes different types of Bayesian and non-Bayesian spatiotemporal models.

Bayesian Model Selection under Constraints (BMSCS)

BMSC is a Bayesian model selection algorithm used for causal analysis and other applications.

Bayesian Prediction (Bpred)

Bayesian multivariate regression application.

ReSources

Bayesian mixing model.

PlotR

PlotR produces Bayesian smoothed plots.

OsteoBioR

OsteoBioR is used to model sequential data where each measurement is known to cover a time interval (integration time). This can be used, among others, to model isotopic temporal trajectories from isotopic measurements taken in multiple living tissues.

TraceR

TraceR helps to generate a visual representation of a relational map.

MapR

MapR displays temporal and temperature graphical files for Isomemo

CausalR

CausalR is a Wrapper for the CausalImpact Package - detect changes in time series trends following different link functions

BNR Project

This project leverages graphical models to perform queries with constrained bayesian estimation. This work is developed as a Shiny app in order to provide an easy and clean interaction.

© 2023 MPI-GEA | Pandora & IsoMemo Comutational Lab