Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
414
datasets available to search
ShareScore release 0.9.0
Dataset results
414 results for “Generative Model”
Code generation for classical-quantum software systems modelled in UML - Dataset and EGL Transformation
<p>This dataset contains all the elements necessary for carry out the EGL transformation from UML models to Hybrid and Quantum code, as well as to carry its validation. </p> <blockquote> <p><em>Quantum computing is gaining an increasing interest since it can solve certain problems exponentially faster than classical computing. Thus, many organizations are researching and launching investments for integrating quantum software into their existing systems. Software modernization (as based on Model-Driven Engineering) has been proposed to migrate from/to the so-called hybrid software systems, which integrate classical and quantum software. In that process, both, reverse engineering and restructuring phases, have already been investigated. However, forward engineering phase for generating hybrid source code from high-level design models has not yet been addressed. Thus, this research proposes a quantum code generation technique from extended UML design models. It consists of a set of Model-to-Text transformations (defined through Epsilon Generation Language) to generate both Python and Qiskit code, which respectively integrate classical and quantum code. The transformation has been validated through a multi-case study with 7 hybrid software systems modelled in UML, which demonstrated that the transformation is effective and efficient. The implication of this work is that the software modernization process for hybrid software systems can be completed by tackling forward engineering phase, and that Model-Driven Engineering can therefore globally facilitate industry adoption of quantum software.</em></p> </blockquote>
Model Generation from Requirements with LLMs: an Exploratory Study - Replication Package
<p>This is a replication package for the paper "<span>Model Generation from Requirements </span><span>with LLMs: an Exploratory Study</span>", by Sallam Abualhaija, Chetan Arora, and Alessio Ferrari.</p> <p><strong>Abstract: </strong>Complementing natural language (NL) requirements with graphical models can improve stakeholders’ communication and provide directions for system design. However, creating models from requirements involves manual effort. The advent of generative large language models (LLMs), ChatGPT being a notable example, offers promising avenues for automated assistance in model generation. This paper investigates the reliability of ChatGPT in generating sequence diagrams from NL requirements. Specifically, we conduct a qualitative study examining the sequence diagrams generated by ChatGPT for 28 requirements documents of various types and from different domains. Our study aims to uncover potential issues that emerge in the models generated by ChatGPT, thereby hindering its applicability in practice. Observations have systematically been captured through evaluation logs, and categorized through thematic analysis. Our results indicate that, although the models generally conform to the standard and exhibit a reasonable level of understandability, their correctness with respect to the specified requirements often presents challenges. This issue is particularly pronounced in the presence of requirements smells, such as ambiguity and inconsistency. The insights derived from this study can influence the practical utilization of LLMs in the RE process, and open the door to novel RE-specific prompting strategies targeting effective model generation.</p> <p>The replication package consists of the following folders:</p> <p><strong>logs:</strong> includes the evaluation logs produced by each evaluator</p> <p><strong>original-documents: </strong>includes the original requirements documents used for the evaluation</p> <p><strong>RQ1 - quantitative analysis:</strong> includes the analysis made on the scores given to each model and model variant. It includes five files:</p> <p>- results.csv: numerical results of the evaluation for each criterion<br>- analysis-results.Rmd: R file used to perform the quantitative analysis (requires R Studio to be executed)<br>- analysis-results.html: html file produced by analysis-results.Rmd<br>- cross-check.csv: file with the cross-checking of the two assessors applied to a subset of the models<br>- symmary_results.xlsx: final output of the quantitative results in terms of Wilcoxon signed rank tests</p> <p><strong>RQ2 - thematic analysis: </strong>includes the codebook produced by the thematic analysis of the issues in generating models with ChatGPT</p>
Datasets generated in the ConvAE-RF modelling of grain yield in the mid-lower Yangtze plains
<p><em>formatted_grid.zip </em>is the preprocessed input meteorological dataset to a ConvAE-RF model proposed by the author.</p> <p><em>final_output.zip</em> is 2016-2100 output yield projections from ScenarioMIP experiments SSP126, SSP245, SSP370 and SSP585 of 25 AMES NEX GDDP CMIP6 GCMs downscaled by <a title="NASA Global Daily Downscaled Projections, CMIP6" href="https://www.nature.com/articles/s41597-022-01393-4" target="_blank" rel="noopener">Thrasher et al., 2022.</a></p> <p><em>coldwave_order.csv</em> contains 4 lists of 25 AMES NEX GDDP CMIP6 GCMs (one for each ScenarioMIP experiment) ranked according to mean coldwave frequency predicted in a unit area (0.25° * 0.25°) in the mid-lower Yangtze plain provinces.</p> <p><em>heatwave_order.csv</em> contains 4 lists of 25 AMES NEX GDDP CMIP6 GCMs (one for each ScenarioMIP experiment) ranked according to mean heatwave frequency predicted in a unit area (0.25° * 0.25°) in the mid-lower Yangtze plain provinces.</p> <p><em>tmax_thr90.nc</em> is 2d and records the threshold daily maximum temperature for each point on the spatial grid above which a day would be qualified as a heatwave candidate.</p> <p><em>tmax_thr90.nc</em> is 2d and records the threshold daily minimum temperature for each point on the spatial grid below which a day would be qualified as a coldwave candidate.</p> <p>Please reference the README in this <a href="https://github.com/zjmagou/MLYPGrain2024">GitHub Repository</a> for data usage.</p> <p> </p>
Data and Software of "Development of a Geometric Modeling Strategy for the Generation of Representative Unit Cells in 2D Braids"
<h1><strong>Id: Data of following publication</strong></h1> <p>title = "Development of a Geometric Modeling Strategy for the Generation of Representative Unit Cells in 2D Braids",<br>journal = "<span>Composite Structures</span>",<br>volume =" 348",<br>pages = "118503",<br>year = "2025",<br>doi = "<a href="https://doi.org/10.1016/j.compstruct.2024.118503" target="_blank" rel="noopener">10.1016/j.compstruct.2024.118503</a>",<br>author = "José Rothkegel, Benjamin Renson, Michael Bruyneel, Ludovic Noels"</p> <p>Data doi on 10.5281/zenodo.10829042</p> <h1>pyRVE</h1> <h2><em>Python Code for Geometrical Generator for Braided Composites RVE</em></h2> <p>pyRVE is a code written in <em>Python</em> using the <em>GMSH API</em> that generates the Representative Unit Cell (RUC) of braided composites. It allows the generation of the RUC of triaxial braided for <em>Diamond</em> and <em>Regular</em> patterns.</p> <h2>Requirements</h2> <p>To run, it requires:</p> <ul> <li>The GMSH Python API, which must be built with OpenCascade support. <ul> <li>Choose a local installation directory; <code>CMAKE_INSTALL_PREFIX=$HOME/local/gmsh</code>, and <code>GMSHPY_INSTALL_DIRECTORY=$HOME/local/gmsh</code> e.g.;</li> <li>Make that directory part of your <code>export PYTHONPATH=$HOME/local/gmsh/lib:$PYTHONPATH</code>.</li> <li>After compiling use <code>make install</code>.</li> </ul> </li> <li>The CM3 app dG3D if the final RVE homogenized solution is needed (<a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a>).</li> <li>Make sure that the latest version of OpenCascade (OCCT) is used. Current used version in occt-V7.8.0.</li> </ul> <h2>Usage</h2> <h3>File Structure</h3> <p>A typical run case must have a file structure, where:</p> <ul> <li><code>brd</code>: the files <code>.brd</code> and <code>.brep</code> are located here. The <code>.brd</code> is a backup of the <code>braidClass</code> instance used in the model saved using <code>pickle</code>, the <code>.brep</code> is the Boundary Representation file that can be opened with <em>GMSH</em>.</li> <li><code>csv</code>: the <code>.csv</code> file saved here is the initial output of the code. It contains the actually used dimensions and the final cover factor of the braid.</li> <li><code>data</code>: It contains <code>.csv</code> files with the material properties and the dimensions of the tows. The original model dimensions are read from here.</li> <li><code>dir</code>: In the case of running the RVE homogenization, the directions of the tow fibers are stored here. They are saved for post processing.</li> <li><code>msh</code>: the mesh file <code>.msh</code> obtained after the geometry geneartion is stores here.</li> <li><code>png</code>: in the case of automatic post processing, png files are stored here.</li> <li><code>res</code>: this folder is used to store the homogenization results. They have to be moved here.</li> <li><code>stp</code>: if acitvated, a <code>.stp</code> file of the geometry is stored here</li> <li><code>svg</code>: the projection of the geometry on the <em>x-y</em> plane is stored here.</li> <li><code>vtk</code>: A copy of the mesh file without the matrix mesh is sotred here as a `.vtk`` file.</li> </ul> <h3>How to Run</h3> <p>We will consider the current file structure to run the example in 000_Base. To run the code, it can be called from the command prompt as</p> <div> <pre><code>python3 ../../source/mainRVE.py --name <i> --pattern <pattern></code></pre> </div> <p>In this case, the <code>--name</code> refers to the index that will be given to the model, where <code><i></code> must be changed to an integer and <code>--pattern</code> refers to the wanted pattern to be used, where <code><pattern></code> must be changed to either <code>dia</code> or <code>reg</code>.</p> <blockquote> <p>Note: <code><code>--name</code>cat</code> can also be used to reproduce the regular pattern benchmark of the paper. In that case, the volume fraction of fiber in the tows is hard coded as the provided value in the reference (i.e. 0.86). For other cases, the volume fraction is evaluated from the tow cross-sections.</p> <p>Note: <code>mainRVE.py</code> must be accesible from the directory where the case is being run. This example shows the usage of the current file structure.</p> </blockquote> <h3>All Command Line Options</h3> <p>The code can be run using further options that serve different purpouses, some serving pre processing needs and other serving run administration. The different command line options are:</p> <ul> <li>Required: <ul> <li><code>--name</code> : it gives a suffix to the run model. It is usually an integer.</li> <li><code>--pattern</code> : indicates the type of pattern to be used to build the geometry. The two current options are <code>dia</code> for diamond and <code>reg</code> for regular.</li> </ul> </li> <li>Optional <ul> <li><code>-dG3D</code>: it indicates that the homogenization of the generated RUC is to be perfomed.</li> <li><code>-GMSH</code> : it indicates that GMSH must be open upon competion of the generation of the mesh.</li> <li><code>-loadModel</code> : it will try to load a premade model. It will ignore <code>--pattern</code>.</li> <li><code>--rndPrm</code> : it will generate randomized geometrical parameters. It can be used to generate batches of results. It takes an argument that can be <code>2</code>, <code>4</code> or <code>6</code>. Currently, <code>2</code> gives a random value for <code>s_axial</code> and <code>theta</code>, <code>4</code> randomizes the same as <code>2</code> and adds <code>h_axial</code> and <code>h_bias</code>, and <code>6</code> randomizes the same as <code>4</code> and adds <code>w_axial</code> and <code>w_bias</code>.</li> </ul> </li> <li>Pre-Processing <ul> <li><code>-refCF</code>: it tells the code to generate a grid of values for <code>s_axial</code> and <code>theta</code> where only the cover factor is obtained. It is meant for posterior graphing purposes.</li> </ul> </li> </ul> <h3>Examples</h3> <p>Following the run options, a few examples are indicated</p> <ul> <li>A basic mesh generation run for the basic data, considering a <strong>regular pattern</strong>, for a model named <strong>2</strong>:</li> </ul> <div> <pre><code>python3 ../../source/mainRVE.py --name 2 --pattern reg</code></pre> </div> <ul> <li>The generation of the cover factor data and export, considering a <strong>regular pattern</strong>:</li> </ul> <div> <pre><code>python3 ../../source/mainRVE.py --pattern reg -refCF</code></pre> </div> <ul> <li>A run for the modified basic data, where the <strong>2</strong> parameters are modified <em>randomly</em>, considering a <strong>regular pattern</strong>, for a model named <strong>2</strong>:</li> </ul> <div> <pre><code>python3 ../../source/mainRVE.py --name 2 --pattern reg --rndPrm 2</code></pre> </div> <ul> <li>A run, where model <strong>2</strong> already exists in <code>brd</code> folder but not the <code>.msh</code> and <code>.vtk</code> files:</li> </ul> <div> <pre><code>python3 ../../source/mainRVE.py --name 2 -loadModel </code></pre> </div> <h2>Code Structure</h2> <p>The code is implemented into Python files, where <code>mainRVE.py</code> runs the whole code. The files are:</p> <ul> <li>Braid: <ul> <li><code>braidClass.py</code> :</li> <li><code>bzrPairClass.py</code> :</li> </ul> </li> <li>Geometry <ul> <li><code>bezrClass.py</code> :</li> <li><code>bilnClass.py</code> :</li> <li><code>patchClass.py</code> :</li> <li><code>pntSetClass.py</code> :</li> <li><code>pointClass.py</code> :</li> <li><code>sctnClass.py</code> :</li> <li><code>stripeClass.py</code> :</li> <li><code>surfClass.py</code> :</li> <li><code>surfOffClass.py</code> :</li> </ul> </li> <li>Material: <ul> <li><code>chamis.py</code> :</li> </ul> </li> <li>Tools: <ul> <li><code>dataIO.py</code> :</li> <li><code>postDirection.py</code> :</li> <li><code>tool.py</code> :</li> <li><code>toolData.py</code> :</li> </ul> </li> <li><code>curveClass.py</code> :*</li> </ul> <p> </p> <p> </p>
Predicting transcriptional responses to novel chemical perturbations using deep generative model for drug discovery
<p>Understanding transcriptional responses to chemical perturbations is central to drug discovery, but exhaustive experimental screening of diseasecompound combinations is unfeasible. To overcome this limitation, here we introduce PRnet, a perturbation-conditioned deep generative model that predicts transcriptional responses to novel chemical perturbations that have never experimentally perturbed at bulk and single-cell levels. Evaluations indicate that PRnet outperforms alternative methods in predicting responses across novel compounds, pathways, and cell lines. PRnet enables gene-level response interpretation and in-silico drug screening for diseases based on gene signatures. PRnet further identifies and experimentally validates novel compound candidates against small cell lung cancer and colorectal cancer. Lastly, PRnet generates a large-scale integration atlas of perturbation profiles, covering 88 cell lines, 52 tissues, and various compound libraries. PRnet provides a robust and scalable candidate recommendation workflow and successfully recommends drug candidates for 233 diseases. Overall, PRnet is an effective and valuable tool for gene-based therapeutics screening.</p>
S3GM: Learning spatiotemporal dynamics with a pretrained generative model
<h1>Datasets of Kuramoto-Sivashinsky equation (KSE) and Kolmogorov flow.</h1> <h2>Description of KSE data:</h2> <p>Each file for KSE datasets contains 4 dimensions in the following order: (B*V)*T*X*C. Details are listed in the following table:</p> <table> <tbody> <tr> <td>B</td> <td>number of varying initial conditions</td> </tr> <tr> <td>V</td> <td>number of varying parameters</td> </tr> <tr> <td>T</td> <td>number of temporal frames</td> </tr> <tr> <td>X</td> <td>spatial resolution</td> </tr> <tr> <td>C</td> <td>number of variables in solution (C = 1 for KSE)</td> </tr> <tr> <td>values of parameter used to generate <strong>training </strong>dataset</td> <td>1.0, 1.2, 1.4, 1.6, 1.8, 2.0, 2.2, 2.4, 2.6, 2.8, 3.0, 3.2, 3.4, 3.6, 3.8, 4.0, 4.2, 4.4, 4.6, 4.8, 5.0</td> </tr> <tr> <td>values of parameter used to generate <strong>test </strong>dataset</td> <td>1.1, 2.5, 3.2</td> </tr> </tbody> </table> <h2>Description of Kolmogorov flow data:</h2> <p>Each file for Kolmogorov flow contains 5 dimensions inthe following order: (B*Re*K)*T*X*X*C. Details are listed in the following table:</p> <table> <tbody> <tr> <td>B</td> <td>number of varying initial conditions</td> </tr> <tr> <td>Re</td> <td>number of varying Reynolds numbers</td> </tr> <tr> <td>K</td> <td>number of varying source terms (controled by the value of k)</td> </tr> <tr> <td>T</td> <td>number of temporal frames</td> </tr> <tr> <td>X</td> <td>spatial resolution</td> </tr> <tr> <td>C</td> <td>number of variables in solution (C = 2 for Kolmogorov flow)</td> </tr> <tr> <td>values of Reynolds number used to generate <strong>training </strong>dataset</td> <td>100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1050</td> </tr> <tr> <td>values of Reynolds number used to generate <strong>test </strong>dataset</td> <td> <div> <div>50, 125, 575, 1100, 1500</div> </div> </td> </tr> <tr> <td>values of k used to generate <strong>training </strong>dataset</td> <td>2, 3, 4, 5, 6, 7, 8</td> </tr> <tr> <td>values of k used to generate <strong>test </strong>dataset</td> <td> <div> <div>2, 4, 6, 8</div> </div> </td> </tr> </tbody> </table> <h2>Description of ERA5 data:</h2> <p>Training and testing dataset for ERA5 contains 5 dimensions inthe following order: 1*T*X*X*C, which is manually collected from <a href="https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=download">https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=download</a>. <strong>Note that the quantities in the datasets are already rescaled</strong> (the scale factors are saved in the scalar_era5.npy file, which is a 4x2 array recording the means and stds for the 4 quantities we used). Details are listed in the following table:</p> <table> <tbody> <tr> <td>T</td> <td>number of temporal frames</td> </tr> <tr> <td>X</td> <td>spatial resolution</td> </tr> <tr> <td>C</td> <td>number of variables in solution (C = 4 for ERA5)</td> </tr> <tr> <td>time span for <strong>training </strong>dataset</td> <td>1979-2022</td> </tr> <tr> <td>time span for <strong>test </strong>dataset</td> <td> <div> <div>2023</div> </div> </td> </tr> </tbody> </table> <h2>Pretrained checkpoints:</h2> <p>The .zip file contains the pretrained checkpoints for KSE, Kolmogorov flow and ERA5. Within the .zip file, the folder'kse_v0' is the checkpoint for KSE, 'kol_v0' is the checkpoint for Kolmogorov flow, and 'era5_v0' is the checkpoint for ERA5.</p> <h1><em>Source code:</em></h1> <p>The source code is upload as Github repository in <a href="https://github.com/lzy12301/S3GM">https://github.com/lzy12301/S3GM</a></p>
Code and data for Porting the Meso-NH Atmospheric Model on Different GPU Architectures for the Next Generation of Supercomputers (version MESONH-v55-OpenACC)
<p>GeometricMG.pdf (source: https://bitbucket.org/em459/tensorproductmultigrid/src/master/Documentation/)<br>MESONH_Bench_HECTOR_ADASTRA_LEONARDO.tar.gz: code and data for Meso-NH bench<br>Performance.zip: code and data for figures related to performance<br>WeatherApplications.zip: namelists for running weather applications<br>OASIS3_WW3.tar.gz: OASIS and WW3 codes for running the Meso-NH WWW3 coupled simulation</p>
Test models and test results to evaluate CAD assembly modules capabilities to generate component interfaces
<p>Set of 3D CAD assembly models in STEP AP 203 format.</p> <p>Assembly test models are devoted to evaluations of interfaces between components. The interfaces can be of type surface, rectilinear contacts, circular contacts, or point contacts.</p> <p>Test results obtained from some commercially available CAD assembly modules are given as a set of tables organized in accordancce with contact categories (surface, rectilinear, circular, point).</p> <p>The content and use of the test models are described into the pdf document: Test models and test results to evaluate CAD assembly modules capabilities to generate component interfaces.</p> <p> </p>
Finding Efficient Trade-offs in Multi-Fidelity Response Surface Modeling: Generated data files and figures
<p>All data files and figures generated for the paper "Finding Efficient Trade-offs in Multi-Fidelity Response Surface Modeling".</p> <p>The code used to generate this is archived at <a href="https://doi.org/10.5281/zenodo.6123254">zenodo.org/record/6123254</a></p>
Hadron Shower Simulation Data for Generative Models in Fundamental Physics
<p>Data set containing pion calorimeter showers used to train and evaluate our generative models for our Hadrons, Better, Faster, Stronger publication. Complete dataset consist of three hdf5 files:</p> <ul> <li>pion_train_uniform.hdf5 contains showers originating form pions with a uniformly distributed energy ranging form 10 GeV to 100 GeV. This set was used to train the models.</li> <li>pion_eval_uniform.hdf5 contains showers originating form pions with a uniformly distributed energy ranging form 10 GeV to 100 GeV. This set was used to evaluate the models.</li> <li>pion_eval_steps20to90.hdf5 contains showers originating form pions with discrete energies ranging form 20 GeV to 90 GeV in steps of 10 GeV. This set was used to evaluate the models.</li> </ul> <p>Each file contains a group called 'hcal_only'. This group has two dataset, 'energy' which contains the energy of the pions in GeV and 'layers' which contains the shower images, projected onto a 48x48x48 grid. For our model training this was reduced to 48x25x25 via slicing. The entries correspond to energy depositions in MeV.</p>
Deep learning generates custom-made logistic regression models for explaining how breast cancer subtypes are classified
<p>Breast cancer is the most frequently found cancer in women and the one most often subjected to genetic analysis. Nonetheless, it has been causing the largest number of women's cancer-related deaths. PAM50, the intrinsic subtype assay for breast cancer, is beneficial for diagnosis and stratified treatment but does not explain each subtype's mechanism. Nowadays, deep learning can predict the subtypes from genetic information more accurately than conventional statistical methods. However, the previous studies did not directly use deep learning to examine which genes associate with the subtypes. Ours is the first study on a deep-learning approach to reveal the mechanisms embedded in the PAM50-classified subtypes. We developed an explainable deep learning model called a point-wise linear model, which uses a meta-learning approach to generate a custom-made logistic regression model for each sample. Logistic regression is familiar to physicians and medical informatics researchers, and we can use it to analyze which genes are important for subtype prediction. The custom-made logistic regression models generated by the point-wise linear model for each subtype used the specific genes selected in other subtypes compared to the conventional logistic regression model: the overlap ratio is less than twenty percent. And analyzing the point-wise linear model's inner state, we found that the point-wise linear model used genes relevant to the cell cycle-related pathways. The results of this study suggest the potential of our explainable deep learning to play a vital role in cancer treatment.</p>
Bulk NGS/allele quantification - Highly efficient generation of isogenic pluripotent stem cell models using prime editing
<p>This file contains sequencing results for bulk NGS and amplicons covering specific regions of interest related to prime editing of hPSCs. </p>
Highly efficient generation of isogenic pluripotent stem cell models using prime editing - Datasets
<p>This collection contains the following datasets:</p> <p>1. AAVS1 knock-in genotyping</p> <p>2. aCGH karyotyping</p> <p>3. Tabular datasets for associated graphical figures </p> <p> </p>
Dataset for High-resolution mantle flow models reveal importance of plate boundary geometry and slab pull forces on generating tectonic plate motions
<p>This repository contains the plugin and dataset used to setup models in the manuscript: "High-resolution mantle flow models reveal importance of plate boundary geometry and slab pull forces on generating tectonic plate motions".</p> <p>The material model plugin used in our models is in the "plugins" folder. The "models" folder contains the reference input parameter file described in the paper. All other model configurations shown in the paper can be obtained by modifying this parameter file. The input files used to set up the models are in the respective folder. Additionally, the Jupyter notebook used to compute residuals of our models is provided in the "scripts" folder.</p> <p> </p>
Model generated open-closed field topology maps produced by ISSI team "Magnetic Open Flux And Solar Wind Structuring Of Interplanetary Space"
<p>Open and closed magnetic field topologies of the solar corona generated using four PFSS based coronal models (EUHFORIA, WSA, MULTI-VP, PSI-PFSS) and one full MHD coronal model (PSSI-MHD) using two different types of HMI-ADAPT magnetic field maps (with and without Active Regions (AR) added retrospectively). The magnetograms used are accessible here <a href="https://doi.org/10.5281/zenodo.10211762" target="_blank" rel="noopener">https://doi.org/10.5281/zenodo.10211762</a> (Henney, Carl. ‘ADAPT Global Solar Magnetic Maps - 2010 Sep 18-20 (w/ & W/o Farside Active Region Input)’. Zenodo, 28 November 2023).</p> <p>These model outputs were generated by the ISSI team "Magnetic Open Flux And Solar Wind Structuring Of Interplanetary Space" and were used for the paper accessible on Arxiv via this link: https://arxiv.org/abs/2311.04024</p>
ETimeline: An Extensive Timeline Generation Dataset based on Large Language Model
<div> <div>Timeline generation is of great significance for a comprehensive understanding of the development of events over time. Its goal is to organize news chronologically, which helps to identify patterns and trends that may be obscured when viewing news in isolation, making it easier to track the development of stories and understand the interrelationships between key events. Timelines have appeared in many commercial products, but there is a noticeable lack of research in this field in academia, and existing datasets need improvement in terms of effectiveness and scale. We propose the ETimeline dataset, which contains over 13,000 news articles, covering 600 bilingual timelines across 23 news domains. We collected more than 120,000 news articles as a candidate news pool and used the large language model (LLM) Pipeline to enhance performance, ultimately obtaining the ETimeline, and the news pool data will also be provided. This work contributes to the advancement of timeline generation research and supports a wide range of tasks, including topic generation and event relationships. We believe that this dataset will serve as a catalyst for innovative research and bridge the gap between academia and industry in understanding the practical application of technology services.</div> </div>
Unlocking LLM Insights: A Dataset for Automatic Model Card Generation
<p>Language models (LMs) are no longer restricted to the ML community, and instruction-following LMs have led to a rise in autonomous AI agents. As the accessibility of LMs grows, it is imperative that an understanding of their capabilities, intended usage, and development cycle also improves. Model cards are a widespread practice for documenting detailed information about an ML model. To automate model card generation, we introduce a dataset of 500 question-answer pairs for 25 LMs that cover crucial aspects of the model, such as its training configurations, datasets, biases, architecture details, and training resources. We employ annotators to extract the answers from the original paper. Further, we explore the capabilities of LMs in generating model cards by answering questions. We experiment with three configurations: zero-shot generation, retrieval-augmented generation, and fine-tuning on our dataset. The fine-tuned Llama 3 model shows an improvement of 7 points over the retrieval-augmented generation setup. This indicates that our dataset can be used to train models to automatically generate model cards from paper text and reduce the human effort in the model card curation process.</p>
Dataset related to article 'The Collaborative Seismic Earth Model: Generation 2'
<p>This repository holds 23 earthquake simulations conducted using the second generation of the Collaborative Seismic Earth Model (CSEM2, Noe et al., submitted), serving as benchmark simulations for reproducibility. The simulations are performed with the spectral-element wave propagation solver Salvus (Afanasiev et al., 2019) on event-adapted meshes (Thrastarson et al., 2020). Accompanying waveform data from the stations can be downloaded from public FDSN services and accessed via tools such as the ObsPy mass downloader (Krischer et al., 2015). Information about the earthquakes and receivers is provided in the event files. The source-time function, a filtered heaviside for 50 - 160 s, is included in 'stf.h5' for global simulations and for 20 - 160 s for regional simulations, respectively.</p> <p>The event and waveform information to create Figure 4 in the manuscript are the following:</p> <p>a) <a href="../api/records/11047927/draft/files/GCMT_event_LAKE_TANGANYIKA_REGION_Mag_6.0_2017-2-24-0.h5/content" target="_blank" rel="noopener noreferrer">GCMT_event_LAKE_TANGANYIKA_REGION_Mag_6.0_2017-2-24-0.h5</a>, stf 50 - 160 s</p> <p>b) <a href="../api/records/11047927/draft/files/GCMT_event_CENTRAL_CALIFORNIA_Mag_5.6_2020-6-4-1.h5/content" target="_blank" rel="noopener noreferrer">GCMT_event_CENTRAL_CALIFORNIA_Mag_5.6_2020-6-4-1.h5</a>, stf 20 - 160 s</p> <p>c) <a href="../api/records/11047927/draft/files/GCMT_event_CHILE-ARGENTINA_BORDER_REGION_Mag_5.7_2018-6-21-16.h5/content" target="_blank" rel="noopener noreferrer">GCMT_event_CHILE-ARGENTINA_BORDER_REGION_Mag_5.7_2018-6-21-16.h5</a>, stf 20 - 160 s</p> <p>Simulations for b&c are performed on smaller domains and on regular cubed-sphere meshes.</p>
Benchmarking Study of Deep Generative Models for Inverse Polymer Design: Reinforcement Learning
<p>Well-trained models and generation results for reinforcement learning part of <a href="https://github.com/ytl0410/Polymer-Generative-Models-Benchmark">ytl0410/Polymer-Generative-Models-Benchmark: Well-trained models and generative outcomes for the paper "Benchmarking Study of Deep Generative Models for Inverse Polymer Design" (github.com)</a></p>
Models generated by SIMON
<p>Here you can find information about all models generated by SIMON. Models can be downloaded and re-used for predictions.</p> <p>Each dataset is stored in separate folder which contains all the models built for that dataset.</p> <p>Name format is:<br> {modelName}.RData</p> <p>This file contains following information:</p> <p>- All training specific model data: folds, tuning parameters, etc<br> - All predictions made with test dataset<br> - Confusion matrix and all performance measures calculated<br> - Features and their Variable Importance Scores</p> <p><br> Here is an example of RData file structure:</p> <pre><code class="language-markdown">List of 5 $ model_training_fit :List of 23 ..$ method : chr "bagEarth" ..$ modelInfo :List of 15 .. ..$ label : chr "Bagged MARS" .. ..$ library : chr "earth" .. ..$ type : chr [1:2] "Regression" "Classification" .. ..$ parameters:'data.frame': 2 obs. of 3 variables: .. .. ..$ parameter: Factor w/ 2 levels "degree","nprune": 2 1 .. .. ..$ class : Factor w/ 1 level "numeric": 1 1 .. .. ..$ label : Factor w/ 2 levels "#Terms","Product Degree": 1 2 .. ..$ grid :function (x, y, len = NULL, search = "grid") .. ..$ loop :function (grid) .. ..$ fit :function (x, y, wts, param, lev, last, classProbs, ...) .. ..$ predict :function (modelFit, newdata, submodels = NULL) .. ..$ prob :function (modelFit, newdata, submodels = NULL) .. ..$ predictors:function (x, ...) .. ..$ varImp :function (object, ...) .. ..$ levels :function (x) .. ..$ tags : chr [1:5] "Multivariate Adaptive Regression Splines" "Ensemble Model" "Implicit Feature Selection" "Bagging" ... .. ..$ sort :function (x) .. ..$ oob :function (x) ..$ modelType : chr "Classification" ..$ results :'data.frame': 3 obs. of 24 variables: .. ..$ degree : num [1:3] 1 1 1 .. ..$ nprune : num [1:3] 2 10 18 .. ..$ logLoss : num [1:3] 1.27 1.84 1.66 .. ..$ AUC : num [1:3] 0.694 0.75 0.695 .. ..$ Accuracy : num [1:3] 0.623 0.698 0.657 .. ..$ Kappa : num [1:3] 0.12 0.36 0.262 .. ..$ F1 : num [1:3] 0.46 0.614 0.542 .. ..$ Sensitivity : num [1:3] 0.217 0.589 0.517 .. ..$ Specificity : num [1:3] 0.895 0.765 0.743 .. ..$ Pos_Pred_Value : num [1:3] 0.606 0.655 0.6 .. ..$ Neg_Pred_Value : num [1:3] 0.636 0.76 0.715 .. ..$ Detection_Rate : num [1:3] 0.0864 0.238 0.2098 .. ..$ Balanced_Accuracy : num [1:3] 0.556 0.677 0.63 .. ..$ logLossSD : num [1:3] 0.188 0.693 0.562 .. ..$ AUCSD : num [1:3] 0.19 0.146 0.157 .. ..$ AccuracySD : num [1:3] 0.0922 0.1339 0.1302 .. ..$ KappaSD : num [1:3] 0.217 0.28 0.279 .. ..$ F1SD : num [1:3] 0.099 0.174 0.176 .. ..$ SensitivitySD : num [1:3] 0.204 0.246 0.266 .. ..$ SpecificitySD : num [1:3] 0.12 0.194 0.182 .. ..$ Pos_Pred_ValueSD : num [1:3] 0.369 0.257 0.235 .. ..$ Neg_Pred_ValueSD : num [1:3] 0.0711 0.1264 0.137 .. ..$ Detection_RateSD : num [1:3] 0.0818 0.1114 0.1167 .. ..$ Balanced_AccuracySD: num [1:3] 0.0996 0.1358 0.1406 ..$ pred :'data.frame': 720 obs. of 8 variables: .. ..$ pred : Factor w/ 2 levels "high","low": 2 1 2 2 2 2 1 2 2 2 ... .. ..$ obs : Factor w/ 2 levels "high","low": 1 1 2 2 2 2 1 1 1 2 ... .. ..$ rowIndex: int [1:720] 4 26 34 39 43 47 65 4 26 34 ... .. ..$ high : num [1:720] 0.415 0.822 0.39 0.276 0.135 ... .. ..$ low : num [1:720] 0.585 0.178 0.61 0.724 0.865 ... .. ..$ degree : num [1:720] 1 1 1 1 1 1 1 1 1 1 ... .. ..$ nprune : num [1:720] 18 18 18 18 18 18 18 2 2 2 ... .. ..$ Resample: chr [1:720] "Fold01.Rep1" "Fold01.Rep1" "Fold01.Rep1" "Fold01.Rep1" ... ..$ bestTune :'data.frame': 1 obs. of 2 variables: .. ..$ nprune: num 10 .. ..$ degree: num 1 ..$ call : language train.formula(form = factor(outcome) ~ ., data = training, method = model, trControl = trControl, preProcess = NU| __truncated__ ..$ dots : list() ..$ metric : chr "Accuracy" ..$ control :List of 27 .. ..$ method : chr "repeatedcv" .. ..$ number : num 10 .. ..$ repeats : num 3 .. ..$ search : chr "grid" .. ..$ p : num 0.75 .. ..$ initialWindow : NULL .. ..$ horizon : num 1 .. ..$ fixedWindow : logi TRUE .. ..$ skip : num 0 .. ..$ verboseIter : logi FALSE .. ..$ returnData : logi TRUE .. ..$ returnResamp : chr "final" .. ..$ savePredictions : chr "all" .. ..$ classProbs : logi TRUE .. ..$ summaryFunction :function (data, lev = NULL, model = NULL) .. ..$ selectionFunction: chr "best" .. ..$ preProcOptions :List of 6 .. .. ..$ thresh : num 0.95 .. .. ..$ ICAcomp : num 3 .. .. ..$ k : num 5 .. .. ..$ freqCut : num 19 .. .. ..$ uniqueCut: num 10 .. .. ..$ cutoff : num 0.9 .. ..$ sampling : NULL .. ..$ index :List of 30 .. .. ..$ Fold01.Rep1: int [1:73] 1 2 3 5 6 7 8 9 10 11 ... .. .. ..$ Fold02.Rep1: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold03.Rep1: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold04.Rep1: int [1:71] 1 2 3 4 5 7 8 9 10 11 ... .. .. ..$ Fold05.Rep1: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold06.Rep1: int [1:72] 1 2 4 6 7 8 9 10 11 12 ... .. .. ..$ Fold07.Rep1: int [1:73] 1 3 4 5 6 7 8 9 10 11 ... .. .. ..$ Fold08.Rep1: int [1:71] 1 2 3 4 5 6 7 9 10 11 ... .. .. ..$ Fold09.Rep1: int [1:72] 1 2 3 4 5 6 7 8 10 11 ... .. .. ..$ Fold10.Rep1: int [1:72] 2 3 4 5 6 8 9 12 13 14 ... .. .. ..$ Fold01.Rep2: int [1:72] 1 2 4 5 6 7 8 9 10 11 ... .. .. ..$ Fold02.Rep2: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold03.Rep2: int [1:72] 1 2 3 4 5 6 7 9 10 11 ... .. .. ..$ Fold04.Rep2: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold05.Rep2: int [1:71] 1 2 3 4 5 6 7 8 9 11 ... .. .. ..$ Fold06.Rep2: int [1:71] 1 2 3 5 6 7 8 9 10 11 ... .. .. ..$ Fold07.Rep2: int [1:73] 1 3 4 5 6 8 9 10 11 12 ... .. .. ..$ Fold08.Rep2: int [1:73] 2 3 4 5 6 7 8 9 10 11 ... .. .. ..$ Fold09.Rep2: int [1:72] 1 2 3 4 5 6 7 8 10 12 ... .. .. ..$ Fold10.Rep2: int [1:72] 1 2 3 4 7 8 9 10 11 12 ... .. .. ..$ Fold01.Rep3: int [1:72] 1 3 4 6 7 8 9 10 11 12 ... .. .. ..$ Fold02.Rep3: int [1:73] 1 2 3 4 5 6 7 8 10 11 ... .. .. ..$ Fold03.Rep3: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold04.Rep3: int [1:72] 1 2 3 5 6 7 8 9 10 11 ... .. .. ..$ Fold05.Rep3: int [1:72] 2 3 4 5 6 7 8 9 10 11 ... .. .. ..$ Fold06.Rep3: int [1:72] 1 2 3 4 5 6 7 9 10 12 ... .. .. ..$ Fold07.Rep3: int [1:72] 1 2 3 4 5 6 8 9 10 11 ... .. .. ..$ Fold08.Rep3: int [1:71] 1 2 4 5 7 8 9 10 11 13 ... .. .. ..$ Fold09.Rep3: int [1:72] 1 2 3 4 5 6 7 8 9 10 ... .. .. ..$ Fold10.Rep3: int [1:72] 1 2 3 4 5 6 7 8 9 11 ... .. ..$ indexOut :List of 30 .. .. ..$ Resample01: int [1:7] 4 26 34 39 43 47 65 .. .. ..$ Resample02: int [1:8] 24 28 45 56 64 69 72 78 .. .. ..$ Resample03: int [1:8] 20 23 27 40 50 53 57 66 .. .. ..$ Resample04: int [1:9] 6 21 38 46 49 51 54 67 77 .. .. ..$ Resample05: int [1:8] 14 17 42 48 52 62 76 79 .. .. ..$ Resample06: int [1:8] 3 5 15 18 19 36 37 73 .. .. ..$ Resample07: int [1:7] 2 29 33 58 59 71 80 .. .. ..$ Resample08: int [1:9] 8 13 22 30 31 32 35 61 68 .. .. ..$ Resample09: int [1:8] 9 12 44 55 60 70 74 75 .. .. ..$ Resample10: int [1:8] 1 7 10 11 16 25 41 63 .. .. ..$ Resample11: int [1:8] 3 24 27 28 39 53 55 77 .. .. ..$ Resample12: int [1:8] 14 16 36 41 46 59 69 73 .. .. ..$ Resample13: int [1:8] 8 17 31 50 63 70 71 80 .. .. ..$ Resample14: int [1:8] 19 25 35 52 54 58 65 72 .. .. ..$ Resample15: int [1:9] 10 12 13 23 32 38 48 76 78 .. .. ..$ Resample16: int [1:9] 4 21 22 33 34 44 64 67 75 .. .. ..$ Resample17: int [1:7] 2 7 42 49 51 60 79 .. .. ..$ Resample18: int [1:7] 1 15 26 29 37 40 57 .. .. ..$ Resample19: int [1:8] 9 11 18 45 47 56 62 66 .. .. ..$ Resample20: int [1:8] 5 6 20 30 43 61 68 74 .. .. ..$ Resample21: int [1:8] 2 5 34 38 49 53 54 74 .. .. ..$ Resample22: int [1:7] 9 19 26 27 32 70 78 .. .. ..$ Resample23: int [1:8] 17 33 36 46 48 52 64 73 .. .. ..$ Resample24: int [1:8] 4 13 18 21 35 58 63 71 .. .. ..$ Resample25: int [1:8] 1 20 24 28 30 50 55 65 .. .. ..$ Resample26: int [1:8] 8 11 15 22 62 66 72 75 .. .. ..$ Resample27: int [1:8] 7 14 25 31 40 47 59 79 .. .. ..$ Resample28: int [1:9] 3 6 12 42 43 60 69 77 80 .. .. ..$ Resample29: int [1:8] 23 29 41 45 56 57 67 68 .. .. ..$ Resample30: int [1:8] 10 16 37 39 44 51 61 76 .. ..$ indexFinal : NULL .. ..$ timingSamps : num 0 .. ..$ predictionBounds : logi [1:2] FALSE FALSE .. ..$ seeds :List of 31 .. .. ..$ : int [1:9] 114 622 609 999 858 638 10 231 661 .. .. ..$ : int [1:9] 515 693 544 282 920 291 833 285 265 .. .. ..$ : int [1:9] 187 232 316 302 159 40 218 805 522 .. .. ..$ : int [1:9] 915 831 46 455 265 304 505 180 754 .. .. ..$ : int [1:9] 202 259 991 805 552 644 310 618 328 .. .. ..$ : int [1:9] 502 677 485 244 763 74 308 713 501 .. .. ..$ : int [1:9] 153 504 493 749 174 845 860 42 315 .. .. ..$ : int [1:9] 14 239 706 308 507 52 562 121 886 .. .. ..$ : int [1:9] 15 783 90 518 383 70 319 664 919 .. .. ..$ : int [1:9] 472 143 544 196 895 388 310 159 890 .. .. ..$ : int [1:9] 167 900 134 132 105 510 299 27 308 .. .. ..$ : int [1:9] 743 36 564 280 204 134 324 154 129 .. .. ..$ : int [1:9] 436 39 712 101 947 122 219 907 939 .. .. ..$ : int [1:9] 280 124 796 743 913 990 937 483 282 .. .. ..$ : int [1:9] 252 503 496 318 959 631 127 421 908 .. .. ..$ : int [1:9] 468 908 597 630 866 501 978 323 478 .. .. ..$ : int [1:9] 357 627 741 565 977 574 437 227 82 .. .. ..$ : int [1:9] 851 235 987 601 995 374 552 427 572 .. .. ..$ : int [1:9] 433 225 85 636 430 73 798 324 752 .. .. ..$ : int [1:9] 585 709 427 343 757 422 558 116 301 .. .. ..$ : int [1:9] 479 345 600 76 953 23 837 629 308 .. .. ..$ : int [1:9] 743 639 991 128 880 807 817 829 727 .. .. ..$ : int [1:9] 984 639 660 527 317 765 524 728 306 .. .. ..$ : int [1:9] 405 205 984 565 280 185 754 563 925 .. .. ..$ : int [1:9] 639 701 479 848 421 32 257 333 133 .. .. ..$ : int [1:9] 500 802 337 508 493 794 564 106 999 .. .. ..$ : int [1:9] 568 213 749 307 488 985 422 243 216 .. .. ..$ : int [1:9] 690 980 477 772 573 962 793 529 592 .. .. ..$ : int [1:9] 264 280 65 562 262 4 586 517 838 .. .. ..$ : int [1:9] 30 600 268 121 101 745 16 50 742 .. .. ..$ : int 358 .. ..$ adaptive :List of 4 .. .. ..$ min : num 5 .. .. ..$ alpha : num 0.05 .. .. ..$ method : chr "gls" .. .. ..$ complete: logi TRUE .. ..$ trim : logi FALSE .. ..$ allowParallel : logi TRUE ..$ trainingData:'data.frame': 80 obs. of 13 variables: .. ..$ .outcome : Factor w/ 2 levels "high","low": 2 2 1 1 1 1 1 1 1 1 ... .. ..$ CD161_pos_CD45RA_pos_Tregs : num [1:80] 1.68 0.84 0.43 0.56 0.73 0.64 0.53 1.15 0.51 1.38 ... .. ..$ CD27_pos_CD8_pos_T_cells : num [1:80] 85.2 71.9 84.5 83 74.8 66.4 87.7 64.1 87.3 89.5 ... .. ..$ CD85j_pos_CD8_pos_T_cells : num [1:80] 17.7 25.8 17.1 19.1 19.1 28.6 8.31 18.8 11 6.95 ... .. ..$ CD94_pos_CD8_pos_T_cells : num [1:80] 4.31 14.2 3.94 4.48 10.1 25.8 20.3 11 4.16 2.74 ... .. ..$ central_memory_CD8_pos_T_cells: num [1:80] 1.96 3.27 2.77 6.31 7.59 6.02 8.54 5.64 6.36 2.93 ... .. ..$ effector_CD8_pos_T_cells : num [1:80] 14.7 26.9 13.4 11.7 21 18.8 10.6 14.4 6.82 7.72 ... .. ..$ L50_EOTAXIN : num [1:80] -0.14 1.3 0.28 -0.76 0.16 0.4 0.17 0.88 0.73 0.84 ... .. ..$ L50_HGF : num [1:80] -0.06 1.45 -0.14 -1.12 -0.36 0.19 0.1 0.82 1.14 1.35 ... .. ..$ L50_IL7 : num [1:80] -0.11 1.49 -0.1 -0.88 0.07 0.23 0.18 0.99 0.97 1.26 ... .. ..$ L50_MCP3 : num [1:80] -1.38 2 -0.17 0.48 -0.54 1.03 0.8 0.43 1.06 0.77 ... .. ..$ L50_TRAIL : num [1:80] 0.17 1.8 0.21 -1.56 0.34 0.96 0 -0.59 1.43 1.65 ... .. ..$ monocytes : num [1:80] 17.1 12 20.5 21.2 13.4 15.9 18.2 12.7 14 14.6 ... ..$ resample :'data.frame': 30 obs. of 12 variables: .. ..$ logLoss : num [1:30] 2.89 2.01 1.22 3.06 1.56 ... .. ..$ AUC : num [1:30] 0.933 0.867 0.833 1 0.8 ... .. ..$ Accuracy : num [1:30] 0.875 0.625 0.857 0.889 0.625 ... .. ..$ Kappa : num [1:30] 0.714 0.143 0.696 0.769 0.143 ... .. ..$ F1 : num [1:30] 0.8 0.4 0.8 0.857 0.4 ... .. ..$ Sensitivity : num [1:30] 0.667 0.333 0.667 0.75 0.333 ... .. ..$ Specificity : num [1:30] 1 0.8 1 1 0.8 0.6 0.6 1 0.6 1 ... .. ..$ Pos_Pred_Value : num [1:30] 1 0.5 1 1 0.5 ... .. ..$ Neg_Pred_Value : num [1:30] 0.833 0.667 0.8 0.833 0.667 ... .. ..$ Detection_Rate : num [1:30] 0.25 0.125 0.286 0.333 0.125 ... .. ..$ Balanced_Accuracy: num [1:30] 0.833 0.567 0.833 0.875 0.567 ... .. ..$ Resample : chr [1:30] "Fold03.Rep1" "Fold02.Rep1" "Fold01.Rep1" "Fold04.Rep1" ... ..$ resampledCM :'data.frame': 90 obs. of 7 variables: .. ..$ degree : num [1:90] 1 1 1 1 1 1 1 1 1 1 ... .. ..$ nprune : num [1:90] 18 2 10 18 2 10 18 2 10 18 ... .. ..$ cell1 : num [1:90] 2 0 2 1 1 1 1 1 2 2 ... .. ..$ cell2 : num [1:90] 1 3 1 2 2 2 2 2 1 2 ... .. ..$ cell3 : num [1:90] 0 0 0 1 1 1 1 0 0 0 ... .. ..$ cell4 : num [1:90] 4 4 4 4 4 4 4 5 5 5 ... .. ..$ Resample: chr [1:90] "Fold01.Rep1" "Fold01.Rep1" "Fold01.Rep1" "Fold02.Rep1" ... ..$ perfNames : chr [1:11] "logLoss" "AUC" "Accuracy" "Kappa" ... ..$ maximize : logi TRUE ..$ yLimits : NULL ..$ times :List of 3 .. ..$ everything: 'proc_time' Named num [1:5] 2.25 0.56 13.92 156.52 8.29 .. .. ..- attr(*, "names")= chr [1:5] "user.self" "sys.self" "elapsed" "user.child" ... .. ..$ final : 'proc_time' Named num [1:5] 0.776 0.004 0.783 0 0 .. .. ..- attr(*, "names")= chr [1:5] "user.self" "sys.self" "elapsed" "user.child" ... .. ..$ prediction: logi [1:3] NA NA NA ..$ levels : chr [1:2] "high" "low" .. ..- attr(*, "ordered")= logi FALSE ..$ terms :Classes 'terms', 'formula' language factor(outcome) ~ CD161_pos_CD45RA_pos_Tregs + CD27_pos_CD8_pos_T_cells + CD85j_pos_CD8_pos_T_cells + CD94_pos_CD| __truncated__ $ model_prediction :List of 2 ..$ pred_prob:'data.frame': 25 obs. of 2 variables: .. ..$ high: num [1:25] 0.237 0.52 0.625 0.857 0.452 ... .. ..$ low : num [1:25] 0.763 0.48 0.375 0.143 0.548 ... ..$ pred_raw : Factor w/ 2 levels "high","low": 2 1 1 1 2 2 1 1 1 2 ... $ roc_auc :List of 2 ..$ roc_p:List of 15 .. ..$ percent : logi FALSE .. ..$ sensitivities : num [1:26] 1 1 1 0.933 0.867 ... .. ..$ specificities : num [1:26] 0 0.1 0.2 0.2 0.2 0.3 0.4 0.5 0.5 0.5 ... .. ..$ thresholds : num [1:26] -Inf 0.0625 0.0688 0.0753 0.122 ... .. ..$ direction : chr "<" .. ..$ cases : num [1:15] 0.52 0.625 0.452 0.55 0.735 ... .. ..$ controls : num [1:10] 0.237 0.857 0.238 0.354 0.167 ... .. ..$ fun.sesp :function (thresholds, controls, cases, direction) .. ..$ auc : 'auc' num 0.7 .. .. ..- attr(*, "partial.auc")= logi FALSE .. .. ..- attr(*, "percent")= logi FALSE .. .. ..- attr(*, "roc")=List of 15 .. .. .. ..$ percent : logi FALSE .. .. .. ..$ sensitivities : num [1:26] 1 1 1 0.933 0.867 ... .. .. .. ..$ specificities : num [1:26] 0 0.1 0.2 0.2 0.2 0.3 0.4 0.5 0.5 0.5 ... .. .. .. ..$ thresholds : num [1:26] -Inf 0.0625 0.0688 0.0753 0.122 ... .. .. .. ..$ direction : chr "<" .. .. .. ..$ cases : num [1:15] 0.52 0.625 0.452 0.55 0.735 ... .. .. .. ..$ controls : num [1:10] 0.237 0.857 0.238 0.354 0.167 ... .. .. .. ..$ fun.sesp :function (thresholds, controls, cases, direction) .. .. .. ..$ auc : 'auc' num 0.7 .. .. .. .. ..- attr(*, "partial.auc")= logi FALSE .. .. .. .. ..- attr(*, "percent")= logi FALSE .. .. .. .. ..- attr(*, "roc")=List of 8 .. .. .. .. .. ..$ percent : logi FALSE .. .. .. .. .. ..$ sensitivities: num [1:26] 1 1 1 0.933 0.867 ... .. .. .. .. .. ..$ specificities: num [1:26] 0 0.1 0.2 0.2 0.2 0.3 0.4 0.5 0.5 0.5 ... .. .. .. .. .. ..$ thresholds : num [1:26] -Inf 0.0625 0.0688 0.0753 0.122 ... .. .. .. .. .. ..$ direction : chr "<" .. .. .. .. .. ..$ cases : num [1:15] 0.52 0.625 0.452 0.55 0.735 ... .. .. .. .. .. ..$ controls : num [1:10] 0.237 0.857 0.238 0.354 0.167 ... .. .. .. .. .. ..$ fun.sesp :function (thresholds, controls, cases, direction) .. .. .. .. .. ..- attr(*, "class")= chr "roc" .. .. .. ..$ call : language roc.default(response = testing$outcome, predictor = predict_model[, "high"], levels = levels(testing$outcome)) .. .. .. ..$ original.predictor: num [1:25] 0.237 0.52 0.625 0.857 0.452 ... .. .. .. ..$ original.response : Factor w/ 2 levels "high","low": 1 2 2 1 2 1 2 2 2 2 ... .. .. .. ..$ predictor : num [1:25] 0.237 0.52 0.625 0.857 0.452 ... .. .. .. ..$ response : Factor w/ 2 levels "high","low": 1 2 2 1 2 1 2 2 2 2 ... .. .. .. ..$ levels : chr [1:2] "high" "low" .. .. .. ..- attr(*, "class")= chr "roc" .. ..$ call : language roc.default(response = testing$outcome, predictor = predict_model[, "high"], levels = levels(testing$outcome)) .. ..$ original.predictor: num [1:25] 0.237 0.52 0.625 0.857 0.452 ... .. ..$ original.response : Factor w/ 2 levels "high","low": 1 2 2 1 2 1 2 2 2 2 ... .. ..$ predictor : num [1:25] 0.237 0.52 0.625 0.857 0.452 ... .. ..$ response : Factor w/ 2 levels "high","low": 1 2 2 1 2 1 2 2 2 2 ... .. ..$ levels : chr [1:2] "high" "low" .. ..- attr(*, "class")= chr "roc" ..$ auc_p: num 0.7 $ confusion_matrix :List of 6 ..$ positive: chr "high" ..$ table : 'table' int [1:2, 1:2] 7 8 8 2 .. ..- attr(*, "dimnames")=List of 2 .. .. ..$ : chr [1:2] "low" "high" .. .. ..$ reference: chr [1:2] "low" "high" ..$ overall : Named num [1:7] 0.36 -0.333 0.18 0.575 0.6 ... .. ..- attr(*, "names")= chr [1:7] "Accuracy" "Kappa" "AccuracyLower" "AccuracyUpper" ... ..$ byClass : Named num [1:11] 0.2 0.467 0.2 0.467 0.2 ... .. ..- attr(*, "names")= chr [1:11] "Sensitivity" "Specificity" "Pos Pred Value" "Neg Pred Value" ... ..$ mode : chr "sens_spec" ..$ dots : list() ..- attr(*, "class")= chr "confusionMatrix" $ variable_importance:'data.frame': 12 obs. of 4 variables: ..$ score_perc: num [1:12] 100 78.8 63.1 47.6 34.5 ... ..$ features : chr [1:12] "L50_EOTAXIN" "central_memory_CD8_pos_T_cells" "CD94_pos_CD8_pos_T_cells" "L50_TRAIL" ... ..$ rank : int [1:12] 1 2 3 4 5 6 7 8 9 10 ... ..$ score_no : num [1:12] 99.3 78.2 62.7 47.3 34.3 ... </code></pre> <p> </p> <p>File was compressed using 7-Zip available at https://www.7-zip.org/.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.