The proposed Pyp-EEG package is a Python open-source pipeline that makes EEG preprocessing and analysis transparent, modular, and reproducible for practitioners. The pipeline integrates artifact removal using independent component analysis (ICA) with automatic classification via ICLabel, absolute and relative spectral power computation by frequency band and region of interest (ROI), temporal analysis based on fixed windows, and automated execution of basic statistical tests. Additionally, an optional module for directed acyclic network inference using the Multiregression Dynamic Model (MDM) enables effective connectivity learning between brain regions over time, based on the power time series generated.
PhishEvolve is an open-source software package for phishing URL inspection. The software integrates a browser extension, a cloud-hosted blacklist repository containing more than 7000 phishing URLs, and an ONNX-compatible inference service within a cloud–edge architecture. Source code, deployment documentation, software artifacts, and example configurations are publicly available through GitHub and Zenodo. The software has been applied in research on phishing URL detection and browser-assisted cybersecurity systems. PhishEvolve provides software components for URL inspection, blacklist-assisted verification, and ONNX-based URL analysis in cybersecurity research and software development activities.
This paper presents a browser-based software tool for post hoc kinematic analysis of upper-limb rehabilitation exercises performed in immersive virtual reality. The tool was conceived to transform raw telemetry logs into clinically interpretable visualizations and quantitative indicators without requiring ad hoc scripts, laboratory-grade motion-capture systems, or complex installation workflows. CSV files generated during therapy sessions are loaded directly into a client-side web interface, where the software supports temporal segmentation through a region-of-interest selection, interactive trajectory inspection in 2D and 3D, time-series exploration, and automatic computation of movement-quality descriptors such as functional range-of-motion volume, velocity-derived indicators, smoothness-related measures, tortuosity, and head–hand relationship metrics to estimate compensatory behavior. The software consolidates the analytical layer of the Rehab-Immersive ecosystem and supports the interpretation of datasets used in previously published studies on framework-based rehabilitation environments (Herrera et al., 2023) safe bimanual goalkeeper training (Herrera et al., 2024), and adaptive task execution in personalized virtual environments (Herrera et al., 2025). By converting VR telemetry into reusable quantitative evidence, the tool facilitates longitudinal follow-up, improves reproducibility across studies, and lowers the barrier for clinicians and researchers to exploit kinematic data in rehabilitation research.
Interactive multisynchrosqueezing mask decomposition (iMSMD) is a Python-based signal processing tool that employs the multisynchrosqueezing transform (MSST) to generate high-resolution time-frequency representations. Through an intuitive graphical interface, users can interactively select regions of interest in the time-frequency domain and extract specific signal components via inverse MSST reconstruction. This approach effectively addresses mode separation challenges in non-stationary signals while maintaining mathematical rigor through its theoretical foundation in time-frequency wiener filtering.
WOAkMedoids is an R package that applies the Whale Optimization Algorithm (WOA) to k-medoids clustering. The theoretical framework has been published separately, but the original pure-R implementation was too slow for practical use. Version 0.2.2 reimplements performance-critical routines in C++ via Rcpp and RcppParallel, achieving speedups of 100× to 190000× over the pure-R baseline. The package accepts precomputed distance matrices, supports early stopping, and is available on CRAN under the GPL-2+ license.
OpenTAS-Forge is an open-source Python framework for the automated generation, scheduling, and network based simulation of IEEE 802.1Qbv Time-Aware Shaper (TAS) schedules, incorporating IEEE 802.1AS time synchronization for Time-Sensitive Networks (TSN). Developing network topologies and generating TAS schedules is often tedious and error-prone, particularly when factoring-in time synchronization effects. OpenTAS-Forge addresses this challenge by providing a unified, automated toolchain that not only generates and analyzes TAS schedules using Python and Google-OR Tools, but also enables their validation through network-level simulation using OMNeT++. The framework supports different TAS scheduling methods and provides per-stream performance metrics, including end-to-end latency, and jitter.
Variable selection is crucial in high-dimensional biomedical research, yet many methods are unstable and select irrelevant variables. Stability selection improves robustness but may miss important variables and does not offer prediction. To address these limitations, we developed stabel, an R package, a framework that integrates multiple stability selection procedures with diverse models and subsampling strategies, followed by ensemble-based prediction. The package supports binary and continuous outcomes, offering LASSO, sparseSVM, and random forest for variable selection, and any prediction algorithm available in the SuperLearner package. stabel provides a flexible and reliable tool for high-dimensional variable selection and prediction.
Kinematic analysis is a foundation for manipulator optimization. The involved directions include inverse kinematic solving, motion control, and path planning. This paper presents a general kinematic analysis tool (KAT) based on OpenGL. With an intuitive user interface, KAT allows users to control their manipulator by mouse operation. Through the simple system setting, different simulations can be carried out in KAT, such as forward and inverse kinematic solving, workspace visualization, and joint variation analysis. KAT is a versatile and accessible platform for scientific research on industrial robotic manipulators.
The representation of multidimensional phenomena requires clustering methods that are both reliable and interpretable. Traditional k-means, however, faces persistent difficulties in selecting the number of clusters and ensuring adequate cohesion and separation, which may compromise substantive interpretation. To address these limitations, this paper proposes a methodological framework that combines a user-defined interval for the number of clusters, a systematic search over parameter configurations, and iterative sub-indicator selection based on either exhaustive optimization or Shannon entropy, retaining only solutions that satisfy predefined quality thresholds. The study’s applied contribution is Smart k-means, an open-source Python implementation that operationalizes this framework and enables users to explore admissible clustering solutions and visualize the resulting groups. The proposed approach supports the identification of cluster structures with improved internal coherence, external separation, and interpretability in applications involving multidimensional phenomena.
This work presents Top-6-Class MetaStudio (Top6Meta), a Python-based framework for the unified modeling of architected materials and structures. The software package integrates three primary architectural families — strut-based lattices, TPMS, and spinodal stochastic architectures — together with three derivative classes formed through secondary operations, namely hybrid multimorphology designs, functionally graded architectures, and interpenetrating phase composites (IPCs). All classes are implemented within a common modeling paradigm, enabling seamless transitions between shapes and topologies. Top6Meta supports the generation of standalone metamaterials, as well as architected beams, plates, and sandwich structures, bridging mesoscale topology designs with macroscale structures, while supporting both graphical and programmatic workflows.
This paper presents an open-source software framework for automated scoring of paper-based multiple-choice exams. The proposed framework processes scanned or photographed answer sheets, then corrects distortions, extracts student data, and recognizes selections with high accuracy. The framework can also generate structured outputs and annotated images for easy integration into learning platforms. Evaluations on real university exams show that the framework achieves 50 sheets per minute with the error rate below 0.5% on commodity hardware, thereby enabling cost-effective assessment at scale.
The aridagri package provides a unified R framework for comprehensive statistical analysis of agricultural field experiments. It implements analysis of variance (ANOVA) for 16 experimental designs, including all split-plot variations with appropriate error terms, seven post-hoc comparison tests with letter groupings, seven stability analysis methods with integrated genotype ranking, thermal indices, crop growth analysis, path analysis, and structural equation modelling. All 33 exported functions produce publication-ready output with standard error (SE), critical difference (CD), and coefficient of variation (CV). A unified verbose parameter enables interactive and programmatic use. The package is distributed under GPL-3 and available from CRAN and GitHub.
Unsupervised structure discovery text remains a central challenge in data mining, as most clustering algorithms require specifying the number of groups in advance or rely on density parameters difficult to tune. This paper presents mstknnclust, an R package implementing the MST-kNN algorithm, which partitions data by intersecting the Minimum Spanning Tree and the k-Nearest Neighbour graph. The value of k is determined automatically, requiring no user-specified hyperparameters. Benchmarking against well-known algorithms yields an Adjusted Rand Index of 0.65 without manual tuning, outperforming all parameter-free methods. The package is validated across computational linguistics, genomics, and seismology, and is freely available on CRAN.
This work introduces GFDFlow, an object-oriented Python package that implements the Generalized Finite Differences Method (GFDM) for solving two-dimensional transport problems. The Generalized Finite Differences Method (GFDM) is well suited for irregular regions — offering greater flexibility with node distribution and managing irregular geometries without structured meshes. The implementation supports multiple boundary conditions (Dirichlet and Neumann), enabling users to tailor problems to their specific needs. GFDFlow serves as a practical tool for researchers and engineers working on transport phenomena applications by offering an accessible platform for tackling complex 2D flow problems in fields such as heat and mass transfer, and fluid dynamics. The GFDM implementation delivers accurate numerical results while remaining computationally efficient, making it suitable for both educational and research purposes.
Identifying abnormal behavior in water distribution networks using sensor data is a challenging yet crucial task. A widely used strategy relies on data-driven techniques that apply statistical null-hypothesis tests to assess whether sensor measurements are consistent with normal operating conditions. To address the shortage of open-source tools tailored to this problem, this paper presents SIAM (Statistical Identification of Anomalies via Mahalanobis metrics), a Python-based application for the automated detection of anomalous operational states in water distribution networks. SIAM is an open-source solution built on multivariate statistical analysis of sensor time series, designed to be user-friendly and requiring minimal prior expertise.
We present Keycloak-SSI, an open-source extension that integrates Self-Sovereign Identity (SSI) verification into the post-authentication step of Federated Identity Management flows. The extension enables Keycloak to act as an SSI verifier for attributes, eliminating the need for an internal credential store: the IdP defines required attributes and trusted SSI issuers, while users respond to attribute verification requests directly, maintaining control over credentials. The existing authentication flow remains unchanged. This approach minimizes the data protection burden on the IdP, enhances user data sovereignty, and reduces the necessity for formal agreements between IdP and credential issuers, while maintaining trust through SSI mechanisms.
This article presents InsuranceRAG, an open-source software system designed to support intelligent interaction with health insurance data using Retrieval-Augmented Generation (RAG). The software provides a Flask-based web interface that enables users to submit natural language queries, upload insurance policy documents, and receive context-aware responses. It follows a modular architecture comprising independent components for conversational assistance, policy recommendation, and document retrieval, all coordinated through a centralized backend. Shared RAG services integrate text embedding models, FAISS-based semantic search, and configurable large language model inference. The open-source implementation promotes reproducibility, extensibility, and practical adoption in insurance-focused natural language processing applications.
Triply Periodic Minimal Surface (TPMS) structures hold significant potential in engineering applications. However, they are typically modeled using tessellated meshes, which limit geometric accuracy and editability. This study presents a reverse modeling framework that integrates MATLAB with Computer-Aided Design (CAD) software to generate B-spline-based TPMS models. Within the developed TPMSpro software package, feature curves are automatically extracted, optimized, and exported as CAD-compatible scripts, enabling batch generation of B-spline curves for constructing high-precision TPMS structures. The resulting geometries can be exported in standard CAD formats (e.g., STEP), enhancing surface smoothness, modeling flexibility, and facilitating downstream engineering tasks.
The Iowa Gambling Task (IGT) is widely used to investigate decision-making under uncertainty; however, many digital implementations provide limited configurability and insufficient temporal resolution of behavioral events. We present GoGamblerGo (GGG), an open-source platform for configurable IGT paradigms with high-resolution trial-level behavioral data acquisition. The system records deck selections, gain and loss outcomes, cumulative balance, and event-level timestamps for decision and feedback phases. Data are exported in standardized formats to support reproducible research and downstream analysis. Temporal validation of event-level timestamps and comparison with existing behavioral frameworks are included. GGG supports extended IGT and multimodal electrophysiological studies.
