Contreras-Torres, Flavio F.

1 files · 461 KB · zipdeclared

v0.3.2 (2026-07-09) HARMONSMILE is a cheminformatics toolkit for preparing and harmonizing molecular structure tables across PubChem, ChEMBL, and independent SMILES datasets. It provides deterministic table loading, robust identifier handling, RDKit-based SMILES canonicalization, and an auditable harmonization layer for downstream curation, comparison, and machine learning workflows. HARMONSMILE preserves source SMILES, emits a canonical/isomeric/Kekulé RDKit representation for compatibility and deduplication, and produces a harmonized canonical/isomeric/aromatic SMILES representation using controlled RDKit-native standardization policies.

Passalacqua, Evan

1 files · 413 KB · zipdeclared

SceneSim is a turn-based EMT (Emergency Medical Technician) simulation platform. A deterministic engine owns all patient truth (vitals, timing, deterioration, scoring); an LLM is used only to render that state as natural-language roleplay, never to decide clinical outcomes.

Diehl, Martin · Eisenlohr, Philip · Roongta, Sharan · et al.

2 files · 56 MB · xz, zipdeclared

DAMASK® is a unified multi-physics crystal plasticity simulation package. The solution of continuum mechanical boundary value problems requires a constitutive response that connects deformation and stress at each material point. This problem is solved in DAMASK on the basis of crystal plasticity using a variety of constitutive models and homogenization approaches. However, treating mechanics in isolation is no longer sufficient to study emergent advanced high-strength materials. In these materials, deformation happens interrelated with displacive phase transformation, significant heating, and potential damage evolution. Therefore, DAMASK is capable of handling multi-physics problems. Following a modular approach, additional field equations are solved in a fully coupled way using a staggered approach.

Smith, C. Haden · Fields, Woodrow L. · Gonzalez, Julian · et al.

1 files · 2.2 MB · zipdeclared

Numerics is a free and open-source .NET library providing numerical methods, probability distributions, statistical analysis, and Bayesian inference tools for quantitative risk assessment in water resources engineering.

Fobbe, Sean

5 files · 155 MB · pdf, zipdeclared

Überblick Dieses Skript wertet das amtliche Internetangebot Gesetze im Internet der Bundesrepublik Deutschland vollständig aus und ist die Grundlage des Corpus des Deutschen Bundesrechts (C-DBR) . Alle mit diesem Skript erstellten Datensätze werden dauerhaft kostenlos und urheberrechtsfrei auf Zenodo, dem wissenschaftlichen Archiv des CERN, veröffentlicht. Alle Versionen sind mit einem separaten und langzeit-stabilen (persistenten) Digital Object Identifier (DOI) versehen. Die neueste Version des Datensatzes ist zusätzlich immer über den Link der Concept DOI erreichbar: https://doi.org/10.5281/zenodo.3832111 Lesen Sie bitte zuerst den Compilation Report! Dieser enthält den Source Code für die gesamte Pipeline, relevante Rechenergebnisse, alle Diagramme, Zeitstempel, sowie ein detailliertes und klickbares Inhaltsverzeichnis. Sie werden sich auf diese Weise viel schneller im eigentlichen Source Code zurechtfinden. Aktualisierung Dieser Datensatz wird ca. alle 3 Monate aktualisiert. Benachrichtigungen über neue und aktualisierte Datensätze veröffentliche ich immer zeitnah auf Mastodon unter @seanfobbe@fediscience.org NEU in Version 2026-07-09 Vollständige Aktualisierung der Daten Ausführung der Pipeline in Userland im Container Neues Skript für Zenodo Upload Funktionsweise Primäre Endprodukte des Skripts sind folgende ZIP-Archive: Der volle Datensatz im CSV-Format, unterteilt in Einzelnormen (nur Rechtsakte mit veröffentlichtem Normtext) Die Metadaten aller Einzelnormen im CSV-Format (wie 1, aber ohne Text-Variable) Der volle Datensatz im CSV-Format, unterteilt in Rechtsakte (nur Rechtsakte mit veröffentlichtem Normtext) Die Metadaten aller Rechtsakte im CSV-Format (wie 3, aber ohne Text-Variable) Die Metadaten aller veröffentlichten Rechtsakte, im CSV-Format (unabhängig davon ob Normtext veröffentlicht wurde) Der volle Datensatz im XML-Format, unterteilt in Rechtsakte (Originaldaten von GII) Alle Anlagen zu den XML-Dateien im jeweiligen Original-Format (Originaldaten von GII) Alle Rechtsakte im TXT-Format, unterteilt in Rechtsakte (deutlich reduzierter Umfang an Metadaten) Alle Rechtstexte im PDF-Format, unterteilt in Rechtsakte (deutlich reduzierter Umfang an Metadaten) Alle Rechtstexte im EPUB-Format, unterteilt in Gesetze (deutlich reduzierter Umfang an Metadaten) Alle Analyse-Ergebnisse (Tabellen als CSV, Grafiken als PDF und PNG) Netzwerk-Strukturen (Adjazenzmatrizen, Edgelists, GraphML, und Netzwerk-Diagramme) für alle Rechtsakte (experimentell!) Alle Ergebnisse werden im Ordner `output` abgelegt. Zusätzlich werden für alle ZIP-Archive kryptographische Signaturen (SHA2-256 und SHA3-512) berechnet und in einer CSV-Datei hinterlegt. Systemanforderungen Docker Docker Compose 6 GB Speicherplatz auf Festplatte Multi-core CPU empfohlen (8 cores/16 threads für die Referenzdatensätze). In der Standard-Einstellung wird das Skript vollautomatisch die maximale Anzahl an Rechenkernen/Threads auf dem System zu nutzen. Die Anzahl der verwendeten Kerne kann in der Konfigurationsatei angepasst werden. Wenn die Anzahl Threads auf 1 gesetzt wird, ist die Parallelisierung deaktiviert. Anleitung Schritt 1: Ordner vorbereiten Kopieren Sie bitte den gesamten Source Code in einen leeren Ordner (!), beispielsweise mit: $ git clone https://codeberg.org/seanfobbe/c-dbr Verwenden Sie immer einen separaten und leeren (!) Ordner für die Kompilierung. Die Skripte löschen innerhalb von bestimmten Unterordnern (files/, temp/, analysis/ und output/) alle Dateien die den Datensatz verunreinigen könnten --- aber auch nur dort. Schritt 2: Docker Image erstellen Ein Docker Image stellt ein komplettes Betriebssystem mit der gesamten verwendeten Software automatisch zusammen. Nutzen Sie zur Erstellung des Images einfach: $ bash docker-compose build --pull Schritt 3: Datensatz kompilieren Falls Sie zuvor den Datensatz schon einmal kompiliert haben (ob erfolgreich oder erfolglos), können Sie mit folgendem Befehl alle Arbeitsdaten im Ordner löschen: $ git clean -fdx Den vollständigen Datensatz kompilieren Sie mit folgendem Skript: $ bash docker-run-project.sh Ergebnis Der Datensatz und alle weiteren Ergebnisse sind nun im Ordner output/ abgelegt. Pipeline visualisieren Sie können die Pipeline visualisieren, aber nur nachdem sie die zentrale .Rmd-Datei mindestens einmal gerendert haben: > targets::tar_glimpse() # Nur Datenobjekte > targets::tar_visnetwork() # Alle Objekte Troubleshooting Hilfreiche Befehle, um Fehler zu lokalisieren und zu beheben. > tar_progress() # Zeigt Fortschritt und Fehler an > tar_meta() # Alle Metadaten > tar_meta(fields = "warnings", complete_only = TRUE) # Warnungen > tar_meta(fields = "error", complete_only = TRUE) # Fehlermeldungen > tar_meta(fields = "seconds") # Laufzeit der Targets Weitere Open Access Veröffentlichungen (Fobbe) Website - www.seanfobbe.de Open Data - https://zenodo.org/communities/sean-fobbe-data/ Source Code - https://zenodo.org/communities/sean-fobbe-code/ Volltexte regulärer Publikationen - https://zenodo.org/communities/sean-fobbe-publications/ Urheberrecht Der Source Code und alle von mir bereitgestellten Rohdaten stehen unter der GNU General Public License v3.0 oder später. Kontakt Fehler gefunden? Anregungen? Kommentieren Sie gerne im Issue Tracker oder kontaktieren Sie mich über www.seanfobbe.de

Pierre Aubert · Thibaut Oprinsen

1 files · 4.9 MB · zipdeclared

This project aims to provide a simple a powerfull lecture compilation to generate html web sites

Bravo, Santiago

1 files · 13 MB · zipdeclared

Open-source deforestation and degradation monitoring system for the Chapada do Araripe protected area (APA) in Brazil. Builds monthly per-pixel baselines from Sentinel-2, Landsat, and Harmonized Landsat Sentinel-2 imagery, then detects anomalies in vegetation moisture (NDMI), burn (NBR), and greenness (EVI2) indices via z-score thresholds with drought-adjusted sensitivity. Designed to be replicable in other regions. Source code licensed under AGPL-3.0-or-later. Data products (baselines, alerts, time series) licensed under CC-BY-SA-4.0; see DATA_LICENSE.

Duarte Rocha, Alby · Aljoumani, Basem

1 files · 36 MB · zipdeclared

rSTEMMUS_SCOPE is an open-source R package that provides a reproducible interface for configuring, executing, and post-processing simulations with the STEMMUS-SCOPE model. The package simplifies complex model workflows by automating input preparation, parameter configuration, model execution, result extraction, and visualization while preserving the original MATLAB implementation of STEMMUS-SCOPE. It supports reproducible ecohydrological modelling, facilitates software interoperability, and provides a structured workflow for research, teaching, and model development. Version 1.0 accompanies the SoftwareX publication describing the package architecture, workflow, and reproducible modelling framework. Release notes (Version 1.0) Complete R interface for STEMMUS-SCOPE Automated preprocessing workflow Enhanced parameter configuration Improved post-processing and visualization Reproducible Steglitz example Updated documentation Official software release accompanying the manuscript describing rSTEMMUS_SCOPE. Data availability: The SoilProperty dataset is distributed separately from the software and is downloaded automatically by the package from its archived Zenodo record. The dataset is unchanged for this release and has therefore not been duplicated. It can also be downloaded manually from: https://zenodo.org/api/records/15488066/files/SoilProperty.zip/content

Microsoft AI for Good Lab · Microsoft Pytorch-Wildlife project

63 files · 12 GB · zipdeclared

v0.19.0: fix the orca-cascade mobile descriptor to reference orca-detector-v3-fp16-tflite (was the superseded orca-detector-fp16-tflite id, which the cascade could not load); no other model changed. v0.18.0: TensorRT compatibility metadata. Added [inference.trt] mode to all 55 desktop ONNX manifests: mode=on_demand for TRT-compatible models (served on CUDA by default; a TensorRT engine is built only on explicit warm-up, never auto-built); mode=off for models with an in-graph spectral/DFT front-end (perch-v2, orca-ecotype-dclde2026-v1) and OWL (open-vocabulary transformer). ONNX weights are bit-identical to v0.17.0 - manifest-only metadata update. v0.17.0: add general/encoder dinov3-vitl16 (DINOv3 ViT-L/16, 1024-d image embedding, DINOv3 License); correct MDV5a manifest id (Species_Net_MDV5a -> MDV5a). v0.16.0: add BioCLIP 2 image encoder (imageomics/bioclip-2, OpenCLIP ViT-L/14, 768-d image embedding, MIT) — first general/encoder entry in the zoo. v0.15.0 — ONB-3: 4 AddaxAI TF/Keras classifiers (tropicam-ai ConvNeXt-Base/84, terai-nepal EffNetV2-M/10, tasmanian-vertebrates EffNetV2-S/96, peruvian-amazon-sdzwa EffNet/43); tf2onnx from .h5/.keras, onboard_parity GATE PASS. v0.14.0: added 2 Apache-2.0 RT-DETRv2 MegaDetector detectors (MDV6-apa-rtdetr-c [ResNet-18vd, 3 decoder layers] + MDV6-apa-rtdetr-e [ResNet-101vd, 6 decoder layers]); NMS-free rtdetr_topk contract, onboard_parity GATE PASS (c 0.974 / e 0.968). v0.13.0 : added MegaDetector MIT variants MDV6-mit-yolov9-c and MDV6-mit-yolov9-e (exported from the MIT-licensed WongKinYiu/YOLO framework, PyTorch-Lightning checkpoints), parity-gated in sparrow-engine. v0.11.0: added the 18 AddaxAI PyTorch camera-trap classifiers (ONB-1: EfficientNetV2 / ResNet50 / YOLOv8-cls / DINOv2 / SpeciesNet-FX, incl. the nz-species multi-label classifier). v0.12.0 : added 9 MegaDetector detector variants (MDV5b; MD1000 redwood/spruce/larch/cedar/sorrel; MDV6 yolov9-c/e; MDV6 rtdetr-c) and re-published MDV5a + MDV6-yolov10-c/e with cv2 (INTER_LINEAR) non-antialiased bilinear preprocessing for detector fidelity. All parity-gated in sparrow-engine. v0.11.0: added the 18 AddaxAI PyTorch camera-trap classifiers (ONB-1: EfficientNetV2 / ResNet50 / YOLOv8-cls / DINOv2 / SpeciesNet-FX, incl. the nz-species multi-label classifier). v0.11.0: added the 18 AddaxAI PyTorch camera-trap classifiers (ONB-1: EfficientNetV2 / ResNet50 / YOLOv8-cls / DINOv2 / SpeciesNet-FX, incl. the nz-species multi-label classifier). ONNX (and mobile TFLite/LiteRT) models for the sparrow-engine inference library. 24 models . v0.10.0 restructures the bundle. Per-model ZIPs are named <domain>__<task>__<id>.zip (Zenodo has no real folders; the __ delimiter groups files by domain then task when sorted). Taxonomy: Camera Trap / Acoustics / Overhead × Detector / Classifier. Latest-version-only for dev iterations (superseded orca detector v1/v2 dropped); the MegaDetector family is kept whole (MDV5a, MDV6-yolov10-c/e). Each ZIP unpacks to <id>/ with manifest.toml, the model graph, labels, MODEL_CARD.md and LICENSE.md. Multi-license — see each model's LICENSE.md.

Szoszkiewicz, Łukasz

1 files · 1.1 MB · zipdeclared

Complete harvest of the OHCHR Search Library "Voting" collection — roll-call and summary voting records of the UN Commission on Human Rights (1946–2006) and the Human Rights Council (2006–present) — parsed from customised MARC21 cataloguing into dashboard-ready CSVs, with a self-contained, zero-build interactive dashboard. Coverage: 6,346 resolutions / 80,159 individual country votes (1946–2026); record count matches the collection total exactly, verified by three independent harvest methods.

Agis-Torres, Ángel · Fernandes, Vítor Samuel · Navarro-Dorado, Jorge · et al.

1 files · 281 KB · zipdeclared

emgteach is an open-source Python package that provides a unified PySide6 desktop application for real-time acquisition, offline analysis and maximum voluntary contraction (MVC) normalisation of surface electromyography (sEMG) signals. It is designed for hands-on biopotential acquisition in undergraduate physiology teaching laboratories, and is meant to be set up by instructors with different levels of technical background and used directly by students during a practical session. The application is hardware-agnostic through a common AcquisitionDevice interface and ships with two interchangeable low-cost backends: BITalino (revolution) over Bluetooth, and an Arduino RedBoard Plus + MyoWare 2.0 over USB serial (open firmware included). A single setting switches between them. Main features: Three-tab GUI (Acquisition, Analysis, MVC normalisation) wrapping a reusable, Qt-free analytic core. Bilingual interface (English / Spanish) with automatic start-up language detection and an in-app language switch. Two-channel acquisition (e.g. agonist/antagonist) with a stacked two-channel live view. Automatic contraction-onset detection (baseline + k·SD threshold) stored as EDF+ annotations. Muscle-load analysis (Jonsson APDF): static (P10), median (P50) and peak (P90) %MVC levels, both offline and as a live monitor with warning/danger zones. One-click PDF session and MVC/muscle-load reports. Reliable EDF+ output using a buffered-write pattern that avoids a silent file-corruption artefact during continuous streaming. Reproducible synthetic signals for class assignments and hardware-free continuous integration. Assisted selection of significant fragments, with an editable fragment editor (detection parameters and envelope-filter cut-offs) applied to the analysis. Region-of-interest analysis, CSV export of results, and a live signal-quality check during recording. An ECG signal profile alongside EMG, selectable through a profile registry. emgteach runs on Windows, macOS and Linux with Python 3.10–3.12, is covered by a suite of 216 automated tests, and is released under the GPL-3.0-or-later license. Source code, documentation and issue tracker: https://github.com/aagisto-maker/emgteach

Choupina Andrade Silva, Alex Joaquim

2 files · 125 KB · docx, xlsxdeclared

This dataset comprises a lithogeochemical database generated from an iron ore deposit located in the western portion of the state of Minas Gerais, Brazil. The region is well known for iron ore exploration because it hosts numerous small- to medium-sized high-grade iron ore bodies, representing an important metallogenic district within the Brazilian iron ore province. The database was developed to support lithological classification based on the integration of geological logging and whole-rock geochemistry. Lithological classification was established using macroscopic descriptions of drill core, visual assessment of primary structures and textures, and multi-element geochemical analyses obtained from 22 diamond drill holes (DDH). The deposit was classified into seven principal lithotypes: Ferruginous colluvium, Itabirite, High-grade itabirite, Amphibolite Ferruginous amphibolite, Granite and Gneiss A total of 646 drill core samples were analyzed for whole-rock geochemistry. Chemical analyses were performed by ALS Chemex (Lima, Peru) following standardized quality-controlled analytical procedures. Major and minor elements were determined using the ME-XRF21u analytical package, based on X-ray fluorescence (XRF) fusion spectroscopy, which is specifically optimized for iron-rich geological materials. The principal analytes include FeO, Al₂O₃, As, Ba, CaO, Cl, Co, Cr₂O₃, Cu, Fe, K₂O, MgO, Mn, Na₂O, Ni, P, Pb, S, SiO₂, Sn, Sr, TiO₂, V, Zn, and Zr. Analytical detection limits ranged from 0.001 to 0.01 wt.%, depending on the element. Loss on ignition (LOI; analytical code OA-GRA05x) was determined gravimetrically after furnace ignition at 1000°C. Ferrous iron (FeO) concentrations were measured by volumetric titration (analytical code Fe-VOL05) following acid digestion using a sulfuric acid-hydrofluoric acid (H₂SO₄-HF) mixture. The dataset was compiled to support research in lithological classification, machine learning, mineral exploration, applied geochemistry, and mineral resource characterization. It provides a high-quality benchmark for supervised classification methods, statistical analyses, and multivariate data exploration in iron ore systems.

Ichsan, Mochammad Hannats Hanafi

3 files · 187 MB · zipdeclared

This record archives source-code snapshots and implementation materials for the IntuNav and EdgeNav navigation methods in desktop multi-browser virtual environments. The archive contains three GitHub repository snapshots: (1) Multi-Browser-dekstopVR-1, which provides the Unity-based desktop multi-browser virtual environment; (2) IntuNav, which contains the implementation of the IntuNav navigation algorithm; and (3) EdgeNav, which contains the implementation of the EdgeNav navigation algorithm. The archived files are provided to support transparency, reproducibility, and DOI-based citation of the custom software and algorithmic components reported in the associated manuscript.

Schwarz, Friedrich

1 files · 267 KB · zipdeclared

A Python wrapper for the GWDG SAIA (Scalable AI Accelerator) platform REST API, providing object-oriented and functional access to chat completions, voice transcription/translation, document conversion, ARCANA (RAG / knowledge bases), and an OpenAI SDK compatibility layer.

Paul, J. Joyson · Durairaja, R. · Jayakumar, N. · et al.

1 files · 1.1 MB · pngdeclared

FIGURE 1. Iniistius naevus (TNJFU/TH/FF/21, 142 mm SL) from southeastern India (A) lateral view; (B) radiograph. Scale: 10 mm.

Bollino, Maurizio · Zhang, Yikai

1 files · 6.7 MB · pngdeclared

FIGURE 3. A: Habitat and Citrus host plant of Rhinoscapha degei; B: Specimen of Rhinoscapha degei on twig of its hostplant, in natural position.

Bollino, Maurizio · Zhang, Yikai

1 files · 1.9 MB · pngdeclared

FIGURE 2. A: Rhinoscapha degei, holotype, aedeagus in dorsal view; B: ditto in lateral view; C: sternite IX in ventral view; D: everted endophallus in lateral view; E: everted endophallus of Rh. lagopyga in lateral view.

Bollino, Maurizio · Zhang, Yikai

1 files · 4.8 MB · pngdeclared

FIGURE 1. A: Rhinoscapha degei, holotype, male, dorsal habitus; B: idem, lateral; C: idem, ventral; D: Rhinoscapha degei, paratype, female, dorsal habitus; E: idem, lateral; F: idem, ventral.

Moravec, Jiří · Trýzna, Miloš

1 files · 15 MB · pngdeclared

FIGURES 140–141. Biotope of Pogonostoma (Microstenocera) globulithorax Jeannel, 1946. 140—part of dry deciduous forest passing to gallery forest near Lake Ravelobe in Ankarafantsika National Park; 141—detail of the coarse bark of the large tree the adults of this species dwelled.