DjuraSoftware
From research to working software
Open-source Python packages and the web services behind my research platforms. Each entry lists the code, the stack, how to install it and the paper to cite.
01 / Open source
Stable
djura
Developer, Djura
Scientific Python toolkit for earthquake engineering, bundling the algorithms behind the Djura platform: GCIM-based record selection, hazard consistency, EDP-IM prediction, IM conversion, vulnerability modelling and storey loss functions.
pip install "djura[all]"import djura from djura.data_loader import load_data data = load_data() # ground motion metadata, cached after the first call print(djura.cite(all=True)) # papers behind each submodulePaper: Conversion of Seismic Fragility and Vulnerability Models to Alternative Intensity Measures for Regional Risk Analysis, Earthquake Spectra (2026)
Maintained
Integrated Performance-Based Seismic Design (IPBSD)
Developer & maintainer
Design framework that limits expected annual loss (EAL) while targeting a mean annual frequency of collapse (MAFC).
Paper: Integrating expected loss and collapse risk in performance-based seismic design of structures, Bulletin of Earthquake Engineering (2021)
Archived
Storey loss function generator
Developer & maintainer
Generates storey loss functions relating engineering demand parameters to decision variables. A newer version is part of djura, on PyPI and the Djura platform.
Paper: Story loss functions for seismic design and assessment: Development of tools and application, Earthquake Spectra (2021)
Archived
RC moment-resisting frame modeller
Developer & maintainer
Builds nonlinear 3D building or 2D frame models of reinforced concrete moment-resisting frames.
Stable
xgb-rhomut
Developer & maintainer
XGBoost models predicting strength ratio, ductility and period relationships, trained on nonlinear dynamic analyses of 7,292 single-degree-of-freedom systems.
pip install xgb-rhomutimport xgbrhomut model = xgbrhomut.XGBPredict(im_type="sa_avg", collapse=False) prediction = model.make_prediction( period=1, damping=0.05, hardening_ratio=0.02, ductility=4, dynamic_ductility=3.0 ) # {"median": ..., "dispersion": ...}Paper: Next-generation non-linear and collapse prediction models for short- to long-period systems via machine learning methods, Engineering Structures (2024)
Stable
Risk-aware navigation system
Software developer
Modified A* pathfinding over the plant map, weighting routes by live risk, with Node.js and Python services and a NoSQL database for building and map data.
Paper: Risk-aware navigation in industrial plants at risk of NaTech accidents, International Journal of Disaster Risk Reduction (2023)
Stable
ANN-based generalised ground motion model
Collaborator, with Savvinos Aristeidou
Artificial neural network ground motion model contributed to the OpenQuake Engine for seismic hazard analysis.
Paper: Artificial neural network-based ground motion model for next-generation seismic intensity measures, Soil Dynamics and Earthquake Engineering (2024)
Stable
ANN-based correlation models
Collaborator, with Savvinos Aristeidou
Neural network correlation models for classical and next-generation intensity measures, contributed to the OpenQuake Engine.
Paper: Correlation models for next‐generation amplitude and cumulative intensity measures using artificial neural networks, Earthquake Spectra (2025)
02 / Web services
Stable
Built Environment Data web services
Lead software developer
Django back end and Nuxt front end for uploading, browsing and downloading experimental test data, with a documented public API. Containerised with Docker behind Nginx.
Paper: The Built Environment Data platform for experimental test data in earthquake engineering, Earthquake Engineering & Structural Dynamics (2024)
