SoftwareX最新文献

We are introducing FAIR-MAST, a data management system designed for historical diagnostic data from the Mega Ampere Spherical Tokamak (MAST) fusion experiments. Following the FAIR (findability, accessibility, interoperability, and re-usability) principles, our system aims to address current accessibility issues with data that supports artificial intelligence and machine learning (AI/ML) and advanced data analysis. The system features public APIs with a searchable metadata index and object storage for remote data access. The API integrates a high-performance data analysis stack for scalable data analysis and AI/ML application development. Performance analysis demonstrates a tenfold improvement in data access speed compared to the legacy system, enabling more efficient and comprehensive data exploration. Additionally, our system is designed to be adaptable to other tokamak facilities, such as MAST-Upgrade (MAST-U) and the Joint European Torus (JET), to expedite fusion energy research and promote collaboration.

With the increasing use of GIS software's, historical planimetric data such as orthophotos and old maps represent key data sources to analyze spatio-temporal landscape evolution. However, geometric error inherent to these data are too often overlooked, possibly leading to confusing misinterpretation of measured surficial changes. The user-friendly Python tool 'PickShift', based on a Monte-Carlo approach, addresses this critical issue by quantifying the surficial uncertainty associated with any features digitized from historical planimetric data. This software provides a valuable framework for a more accurate assessment of landscape dynamics and associated uncertainties.

Scientific research faces workflow inefficiencies and reproducibility issues. PyGeoweaver, a Python library, helps manage data pipelines to address these problems. It stands out for its accessibility and ease of use, making it ideal for individual researchers and small teams. Unlike traditional systems that demand extensive technical expertise, it is designed to be beginner-friendly, allowing anyone to work with it without encountering significant technical hurdles. It can help researchers intuitively design workflows, retrieve data, and analyse it efficiently. It also offers compelling benefits for AI use cases, enabling researchers to develop and deploy AI models with ease. Serving as a decentralised hub for workflow management, it ensures discoverability, accessibility, and reusability of scientific workflows, fostering collaboration and research reproducibility. The paper will introduce its technical framework, present successful applications, and share insights from user feedback. Based on our research experiences in the past two years, PyGeoweaver has proven to be a valuable tool, enabling researchers to attain significant research progress with minimal technical complexities.

Two-way tests in independent groups designs are the most frequently used statistical techniques in various fields. In this paper, we present twowaytests package to investigate the effects of two treatments and the associated interaction on the dependent variable. The package offers two-way tests involving two-way ANOVA, parametric bootstrap based generalized test, generalized pivotal quantity based generalized test, two-way ANOVA for trimmed means, two-way ANOVA using M-estimators for location, and two-way ANOVA for medians. The package provides some basic descriptive statistics, pairwise comparisons, variance homogeneity and normality tests, and well-arranged graphical approaches. Furthermore, its web application is freely available at http://www.softmed.hacettepe.edu.tr/twowaytests.

Due to the rapid development of AI, research on creativity support tools that allow humans to foster creativity through collaboration with AI is actively progressing. The concept of an AI-based mixed initiative co-creative system (AI-MICCS), which gives the user the initiative, has emerged. Our research focuses on the fashion domain, where creativity plays an important role, and we present CoCoStyle, an AI-MICCS that can support the creativity of fashion designers. CoCoStyle has three main phases (CoResearch, CoDesign, and CoImprovement) and six sub-phases (Build, Pick, Create, Switch, View and Fit). Fashion Designers can efficiently group a large number of desired images using the attribute detection model and $k$-means algorithm, create unique designs using StyleGAN2, and improve designs from various perspectives using transformer embedding and attribute detection results. We believe that CoCoStyle will effectively support fashion designers in their creative design process by allowing users to take the initiative in controlling the AI.

The new trend in urban stormwater management is to preserve natural hydrological conditions by addressing changes from urbanization. The software Chaak: Urban Hydrological Impact Modeling is a new hydroinformatics tool designed for sustainable modeling, simulation, and analysis of stormwater sewer networks. It emphasizes two key aspects: enhancing planning for Sustainable Urban Drainage Systems (SUDS) to restore natural hydrological conditions and providing an alternative to conventional storm drainage designs. Additionally, it encourages global scientific collaboration to quickly advance the tool's capabilities through structured development and quality control.

The methodology under the term model-based software engineering (MBSE) gained importance already around 20 years ago, after the publication of the Model-Driven Architecture (MDA) initiative by the Object Management Group (OMG). This development methodology continues to evolve, giving rise to recent proposals such as low-code or no-code. Something that has not changed, as recent surveys point out, is the need for powerful testing approaches and tools for these new methodologies. In MBSE, test inputs are models, so it is key to have frameworks for model generation. However, the main shortcomings of existing model-generation frameworks are their performance limitations and the need for domain-specific knowledge, which seriously hampers their industrial adoption. In this paper, we present the Yekta low-code framework that allows to generate models in a simple way through the application of metaheuristic algorithms.

In this contribution, we present clickR, an R package intended for data cleaning following a semi-automatic and supervised procedure. Few packages and commercial software with cleaning capacities are available. In all cases, their functionalities just cover part of the overall data pre-processing and do not follow an integral approach to cleaning up the data. In contrast, clickR brings together all functions needed for correcting the main structural, variable-assignment and typographical errors found in databases and allows researchers to have a strict control on the suggested changes. This is possible because the package creates a data frame that keeps track of all the implemented data modifications. To prove its capacity for detecting and fixing errors, we clean a messy database that exhibits multiple types of errors within date, numeric and factor variables.

Integrating different land uses within a geographic area is essential to urban planning and development. Accurate and fast land use mix (LUM) measurement is necessary for evaluating urban diversity and sustainability. In this paper, we present landusemix, a Python package developed to calculate LUM using two distinct indices: the Entropy Index and the Herfindahl–Hirschman Index (HHI). The landusemix package provides tools for GIS researchers and urban planners to measure the diversity and concentration of land use. Detailed descriptions of the methodologies employed and examples of practical usage are provided. Researchers can use this package to calculate LUM quickly and in bulk, and its results can be easily incorporated into further analysis.

AutoChem is a software package consisting of two modules. The first module is a virtual chemical reactor that can generate reaction products starting from inserted reactants and reactions. The process can be iterated, calculating reactions of the products obtained from the previous steps. To avoid a combinatorial explosion of products, constraints can be inserted. AutoChem module can generate 3D structures of the products obtained as well as a list of all the reactions occurred (chemical network).

This can be useful for the calculation of reaction free energies by using computational chemistry programs like Gaussian and ORCA. Usually, the computational step involves the geometry optimization of all products followed by the calculation of vibrational frequencies of the optimized structures, to assess if a local minimum is reached. Finally, when products free energies are obtained, the calculation of the reactions free energies can be done.

The second module of Autochem, called check, helps to perform the latter step, launching the jobs, collecting products free energies, restarting false minima and calculating the reactions free energies. This module is intended as a general use even outside AutoChem and might be used to perform a large number of free energy calculations with a little effort from the user. This permits to have results at the level of theory needed. AutoChem might also be used for teaching organic chemistry and basic cheminformatics dealing with SMILES and SMARTS by inserting reactants and reactions and analyzing the products obtained.