Find research datasets worth reusing

opencc-zeroMar 2023View details →

Materials Science Optimization Benchmark Dataset for High-dimensional, Multi-objective, Multi-fidelity Optimization of CrabNet Hyperparameters

Benchmarks are an essential driver of progress in scientific disciplines. Ideal benchmarks mimic real-world tasks as closely as possible, where insufficient difficulty or applicability can stunt growth in the field. Benchmarks should also have sufficiently low computational overhead to promote accessibility and repeatability. The goal is then to win a "Turing test" of sorts by creating a surrogate model that is indistinguishable from the ground truth observation (at least within the dataset bounds that were explored), necessitating a large amount of data. In materials science and chemistry, industry-relevant optimization tasks are often hierarchical, noisy, multi-fidelity, multi-objective, high-dimensional, and non-linearly correlated while exhibiting mixed numerical and categorical variables subject to linear and non-linear constraints. To complicate matters, unexpected, failed simulation or experimental regions may be present in the search space. In this study, 173219 quasi-random hyperparameter combinations were generated across 23 hyperparameters and used to train CrabNet on the Matbench experimental band gap dataset. The results were logged to a free-tier shared MongoDB Atlas dataset. This study resulted in a regression dataset mapping hyperparameter combinations (including repeats) to MAE, RMSE, computational runtime, and model size for CrabNet model trained on the Matbench experimental band gap benchmark task1. This dataset is used to create a surrogate model as close as possible to running the actual simulations by incorporating heteroskedastic noise. Failure cases for bad hyperparameter combinations were excluded via careful construction of the hyperparameter search space, and so were not considered as was done in prior work. For the regression dataset, percentile ranks were computed within each of the groups of identical parameter sets to enable capturing heteroskedastic noise. This contrasts with a more traditional approach that imposes a-priori assumptions such as Gaussian noise, e.g., by providing a mean and standard deviation. A similar approach can be applied to other benchmark datasets to bridge the gap between optimization benchmarks with low computational overhead and realistically complex, real-world optimization scenarios.

Validation of highly sensitive method based on UHPLC-ESI-MS/MS for the quantification of progestogens and androgens in plant material.

We prepared a highly sensitive method for the quantification of progestogens and androgens in plant materials. This method is based on UHPLC-ESI-MS/MS. We show here the data used for the method validation. This includes the determination of linearity, recovery, precision, limits of detection and limits of quantification.  The general procedure can be found in the txt or pdf file.  The resulting data are collected in the excel file and can be found in the csv files, additonally.

Supplementary material for "Marker and source-marker reprogramming of Most Permissive Boolean networks and ensembles with BoNesis"

This archive contains supplementary material for the paper <a href="https://doi.org/10.48550/arXiv.2207.13307">Marker and source-marker reprogramming of Most Permissive Boolean networks and ensembles with BoNesis</a> by Loïc Paulevé (2023): <ul> <li>sources of the executable paper in Markdown and Jupyter notebook formats</li> <li>benchmark dataset</li> <li>scripts for converting the executable paper in LaTeX for arXiv deposit</li> </ul>

Supplementary Material 7 including Salamandra salamandra occurrence data, topographic, geological and land cover data and node-based resistances

Supplementary material for the article "Habitat connectivity supports the local abundance of fire salamanders (Salamandra salamandra) but also the spread of Batrachochytrium salamandrivorans" by Bolte et al. (2023) published in Landscape Ecology (DOI: 10.1007/s10980-023-01636-8) This folder comprises a .shp file with fire salamander occurrences, topographic and land cover data (GeoTiff) from the northern Eifel region as well as the R Code used for the statistical analysis of salamander habitat suitability and connectivity.

opencc-by-4.0Jul 2021View details →

BAM reference data: results of ASTM E139 -11 creep tests on a reference material of Nimonic 75 nickel-base alloy

Results of creep tests on a certified reference material at T = 600°C and a tensile creep load of 160 MPa are provided. The raw data are available in ASCII format (*.lis files).  The file "Inhalt_Content_V1.1.pdf" contains further information about the files provided. The evaluated results include the times to reach 2% and 4% creep strain, respectively, and the creep rate after 400 h. The tests were carried out in an accredited test laboratory. The calibrations of all measurands and test and measuring equipment are documented. The calibrations meet the requirements of the test procedure and are metrologically traceable.

dryad40/100

Armed to the teeth: Supplementary material, R code, Rdata, mesh, landmarks

The structure, composition, and shape of teeth have been related to dietary specialization in many vertebrate species, but comparative studies on snakes' teeth are lacking. Yet, snakes have diverse dietary habits that may impact the shape of their teeth. We hypothesize that prey properties, such as hardness and shape, as well as feeding behavior, such as aquatic or arboreal predation, or holding vigorous prey, impose constraints on the evolution of tooth shape in snakes. We compared the morphology of the dentary teeth of 63 species that cover the phylogenetic and dietary diversity of snakes, using 3D geometric morphometrics and linear measurements. Our results show that prey hardness, foraging substrate, and the main foraging mechanical challenge are important drivers of tooth shape, size, and curvature. Overall, long, slender, curved teeth with a thin layer of hard tissue are observed in species that need to maintain a grip on their prey. Short, stout, less curved teeth are associated with species that undergo high or repeated loads. Our study demonstrates the diversity of tooth morphology in snakes and the need to investigate its underlying functional implications to better understand the evolution of teeth in vertebrates.

opencc-zeroMar 2023View details →

Data of "Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator"

General Data of <a href="http://doi.org/10.1016/j.ijsolstr.2023.112470">https://doi.org/10.1016/j.ijsolstr.2023.112470</a> related to MOAMMM project. Data related to the publication (we would be grateful if you could cite the paper in the case in which you are using the data): title = "Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator.", journal = "International Journal of Solids and Structures", year = "2023", volume = "283", pages = "112470", doi = "10.1016/j.ijsolstr.2023.112470", author = "Ling Wu, Cyrielle Anglade, Lucia Cobian, Miguel Monclus, Javier Segurado, Fatma Karayagiz, Ubiratan Freitas, and Ludovic Noels" This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 862015. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them. Description BI code and results of the inference of a pressure-dependent visco-elastic visco-plastic model developed in [NGU16] with a umat implementation in <a href="https://gitlab.uliege.be/moammm/moammmPublic/code/-/tree/main/MaterialModels/FiniteStrain/Finite_VEVP">https://gitlab.uliege.be/moammm/moammmPublic/code/-/tree/main/MaterialModels/FiniteStrain/Finite_VEVP</a>. The BI is described in [WU23] .The experimental results used in the BI are reported in [COB22,COB22b]. To run the BI you need the open source code <a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a> If you use these data or model, we would be grateful if you could cite the related papers. Bibliography <ul> <li>[WU23] L. Wu, C. Anglade, L. Cobian, M. Monclus, J. Segurado, F. Karayagiz, U. Santos Freitas, L. Noels, Bayesian inference of high-dimensional finite-strain visco-elastic-visco-plastic model parameters for additive manufactured polymers and neural network based material parameters generator, International Journal of Solids and Structures (2023) 112470: https://doi.org/10.1016/j.ijsolstr.2023.112470</li> <li>[COB22] L. Cobian, M. Rueda-Ruiz, J.P. Fernandez-Blazquez, V. Martinez, F. Galvez, F. Karayagiz, T. Lück, J. Segurado, M.A. Monclus, Micromechanical characterization of the material response in a PA12-SLS fabricated lattice structure and its correlation with bulk behaviour, Polymer Testing 110 (2022) 107556: https://doi.org/10.1016/j.polymertesting.2022.107556 (in Open access)</li> <li>[COB22b] Data of “. Cobian, M. Rueda-Ruiz, J.P. Fernandez-Blazquez, V. Martinez, F. Galvez, F. Karayagiz, T. Lück, J. Segurado, M.A. Monclus, Micromechanical characterization of the material response in a PA12-SLS fabricated lattice structure and its correlation with bulk behaviour, Polymer Testing 110 (2022) 107556” http://dx.doi.org/10.5281/zenodo.6136935 (in Open access)</li> <li>[NGU16] V. D. Nguyen, F. Lani, T. Pardoen, X. Morelle, L. Noels, A large strain hyperelastic viscoelastic-viscoplastic-damage constitutive model based on a multi-mechanism non-local damage continuum for amorphous glassy polymers. International Journal of Solids and Structures 96 (2016): 192-216; https://dx.doi.org/10.1016/j.ijsolstr.2016.06.008, Open access: https://orbi.uliege.be/handle/2268/197898</li> </ul> Directories All the codes and experimental results are in five directories: <ol> <li>experimentalTests: experimental data, see the README.txt in each subdirectory for details</li> <li>BayesianVE: BI of the visco-elastic parameters <ol> <li>PlotExperimentalCurves: to vizualize the experimental curves and prepare the observations for the BI in the VE range <ol> <li>Loadcase_H.py and Loadcase_V.py read experimental results and create Load_ExpVE_H.dat and Load_ExpVE_V.dat, which keep the experimental observations and loading conditions to perform the BI.</li> <li>PrintDir_H & PrintDir_V subdirectories with the functions called by Loadcase_H.py and Loadcase_V.py</li> <li>Load_ExpVE_H.dat and Load_ExpVE_V.dat created files with the observations and loading conditions to perform the BI</li> </ol> </li> <li>VE_V2Step and VE_H: BI for viscoelastic properties of "V" specimen (VE_V2Step) and "H" specimen (VE_H) <ol> <li>BI_allpos_sequence.py runs the BI using Predict_VETest.py and creates the MCMC_VE_....dat</li> <li>WarmStart = True is used to restart an inference</li> <li>MCMC_VE_....dat in the VE_V2Step and VE_H directories are the BI results</li> <li>When proceeding in two steps in VE_V2Step, a first step generates MCMC_VE_VN8_1st.dat whose posterior is used as prior in the second step to generate MCMC_VE_VN8_2nd.dat</li> </ol> </li> <li>CheckBayRes: to visualize predictions of a BI sample and experimental curves <ol> <li>MCMCRes.py is used to check the numerical predictions of a BI parameter sample (read last sample by default, V or H direction can be selected at line</li> <li>ResKGEmu.py plots the evolution of elastic properties with time</li> <li>uses as input VE_V2Step/MCMC_VE_....dat or VE_H/MCMC_VE_....dat</li> <li>uses local ViscoElasticTest.py, line.geo, line. msh as interface with https://gitlab.onelab.info/cm3/cm3Libraries code</li> <li>uses local functions plotExpLoad_Unload.py, plotExp.py</li> </ol> </li> <li>ViscoElasticTest.py, line.geo, line.msh: interface with <a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a> code used by VE_V2Step and VE_H to call the VEVP model</li> </ol> </li> <li>BayesianVEVP: BI of the visco-elastic and visco-plastic parameters <ol> <li>PlotExperimentalCurves: to vizualize the experimental curves and prepare the observations for the BI in the VE-VP ranges <ol> <li>Loadcase_H.py and Loadcase_V.py read experimental results and create Load_ExpVEVP_H.dat and Load_ExpVEVP_V.dat, which keep the experimental observations and loading conditions to perform BI at the viscoplastic stage.</li> <li>PrintDir_H & PrintDir_V subdirectories with the functions called by Loadcase_H.py and Loadcase_V.py</li> <li>Load_ExpVEVP_H.dat and Load_ExpVEVP_V.dat created files with the observations and loading conditions to perform the BI</li> </ol> </li> <li>VP_V2step and VP_H2step: BI for viscoelastic-viscoplastic properties of "V" specimen (VP_V2Step) and "H" specimen (VP_H2Step) <ol> <li>BI_allpos_sequence.py runs the BI using Predict_VETest.py and creates the MCMC_VP_....dat</li> <li>WarmStart = True is used to restart an inference</li> <li>It starts from the VE prosterior as prior, see point 2, and generates a MCMC_VP_?_1of2Steps.dat (? being H or V)</li> <li>Then using MCMC_VP_?_1of2Steps.dat posterior to get a new prior, it generates MCMC_VP_?_2of2Steps.dat (? being H or V)</li> </ol> </li> <li>CheckBayRes: visualize predictions of a BI sample and experimental curves <ol> <li>MCMCRes.py is used to check the numerical predictions with 3 BI parameter samples ([28000, 45000,70000] by default, V or H direction can be selected at line 12) using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V)</li> <li>plot_hist.py is used to plot histograms of all the inferred parameters using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V)</li> <li>Plot_Prop.py plots joints histograms of the inferred parameters using the samples of BayesianVEVP/VP_?2step/MCMC_VP_?_2of2Steps.dat (? being H or V)</li> <li>ResKGEmu.py plots the evolution of elastic properties with time</li> </ol> </li> <li>VEVPTest.py: interface with https://gitlab.onelab.info/cm3/cm3Libraries code used by VP_V2Step and VP_H2Step to call the VEVP model</li> </ol> </li> <li>RandomParametersGenerator: used to generate the parameters from the BI samples, with the same statistical content <ol> <li>Generator <ol> <li>DataProcess.py: creates normalized data for training from final inferred parameters in ../MCMC_ResData and creates ?_dirNormData (? being H or V)</li> <li>KmeanDataProcess.py: performs clustering for the data of H_dirNormDat and creates H_dirNormData_2cluster (no need for V direction because not bimodal)</li> <li>Gan_V.py and Gan_H.py are used to train the random material parameter generators and create the VDir_Gan or HDir_Gan200_0/HDir_Gan200_1</li> <li>GenerateParameters.py generates random parameters using the Gan files VDir_Gan or HDir_Gan200_0/HDir_Gan200_1 and checks the joint histograms of generated parameters, generated parameters are in V_GenData and H_GenData</li> <li>Ganlib.py is used by the generator</li> </ol> </li> <li>CheckRes <ol> <li>GenDataRes.py is used to check the numerical predictions with the generated parameter samples, see point 4) (using V_GenData and H_GenData).</li> <li>Plot_PropGen.py plots joints histograms of the generated parameters using the samples of V_GenData or H_GenData</li> </ol> </li> </ol> </li> <li>MCMC_ResData:All final data used in the paper (they can substitute the ones used here above) <ol> <li>H_direction and V_direction keep the MCMC random walk results of BI.</li> <li>RandomParameterGenerator keeps results of the generator Paper</li> </ol> </li> </ol> Figures of [WU23] <ul> <li>Fig. 5: From directory BayesianVE/PlotExperimentalCurves, run python3 ./PrintDir_V/plotExp_T.py or ./PrintDir_V/plotExp_C.py or ./PrintDir_V/plotExp_R.py</li> <li>Fig. 7: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = "V" and then with direct = "H" and with Var = [0,1,20,24,28,29,30,31]</li> <li>Fig. 8: BayesianVEVP/CheckBayRes/MCMCRes.py with direct = "V" (requires <a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a> code)</li> <li>Fig. 9: BayesianVEVP/CheckBayRes/MCMCRes.py with direct = "H" (requires<a href="https://gitlab.onelab.info/cm3/cm3Libraries"> https://gitlab.onelab.info/cm3/cm3Libraries</a> code)</li> <li>Fig. 11: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = "V" and then with direct = "H" and with Var = [0,1,20,24,28,29,30,31]</li> <li>Fig. 12: RandomParametersGenerator/CheckRes/GenDataRes.py with direct = "V" (requires <a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a> code)</li> <li>Fig. 13: RandomParametersGenerator/CheckRes/GenDataRes.py with direct = "H" (requires <a href="https://gitlab.onelab.info/cm3/cm3Libraries">https://gitlab.onelab.info/cm3/cm3Libraries</a> code)</li> <li>Fig. 14A: From directory BayesianVE/PlotExperimentalCurves, run python3 ./PrintDir_H/plotExp_T.py or ./PrintDir_H/plotExp_C.py or ./PrintDir_H/plotExp_R.py</li> <li>Fig. 15C: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = "V", Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27]</li> <li>Fig. 16C: BayesainVEVP/CheckBayRes/plot_hist.py with direct = "V"</li> <li>Fig. 17C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = "V"</li> <li>Fig. 18C: BayesianVEVP/CheckBayRes/Plot_Prop.py with direct = "H", Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27]</li> <li>Fig. 19C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = "H"</li> <li>Fig. 20C: BayesianVEVP/CheckBayRes/plot_hist.py with direct = "H"</li> <li>Fig. 21D: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = "V" , Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27]</li> <li>Fig. 22D: RandomParametersGenerator/CheckRes/Plot_PropGen.py with direct = "H" , Var = [2,3,8,9,14,15,18,19] and [20,21,22,23,24,25,26,27]</li> </ul>

Supplementary material S1. Donacia (Protodonacia) bienkowskii Bukejs and Alekseev subgen. et sp. nov., holotype, Nr. 3300.142 [RSKM], X-ray micro-CT volume rendering of the habitus without legs and antennae.

Supplementary material S1 in paper: Bukejs A., Alekseev V.I. & McKellard R.C. First described reed beetle (Chrysomelidae: Donaciinae) from amber inclusion: member of the littoral community in the Eocene Baltic amber forest. Historical Biology.

opencc-by-4.0Jan 2023View details →

Supplementary Material for Sarquis & Sierra 2023

This repository includes the Supplementary Material for the publication "Information content in time series of litter decomposition studies and the transit time of litter in aridlands" by Sarquis and Sierra (Biogeosciences).

Supplementary Material from: Spatial characterization of the rotifer Community in a reservoir in the Northeast

Supplementary Material for the paper: Spatial characterization of the rotifer Community in a reservoir in the Northeast submitted to Acta Liminologica Brasiliensia.

An Alkali-Activated Concrete Dataset for Sustainable Building Materials

<ul> <li>Size: 1630 mixes of alkali-activated concretes </li> <li>Material type: Alkali-activated concretes</li> <li>Features: <ul> <li>Molecular composition of the binder powder</li> <li>Weights of precursor types</li> <li>Aggregate weights</li> <li>Alkali solution chemical composition</li> <li>Amount of additional water/superplasticizer</li> <li>Curing process</li> </ul> </li> <li>Properties: <ul> <li>Compressive strengths of each mix.</li> <li>Other mechanical and structural properties (not for each mix) </li> <li>Corresponding CO2 footprints, calculated using an approach that takes into account constituent CO2 footprint values of the mix's constituents obtained from literature, as well as a linear regression of the CO2 footprint due to oven curing</li> </ul> </li> <li>Use: Valuable for selecting alkali activated concrete mixes for various applications as both compressive strength and CO2 emissions are critical factors. Aimed at facilitating the development of sustainable and environmentally friendly construction materials.</li> </ul>

opencc-by-4.0Feb 2023View details →

Supporting material for "Scientists should not increase the overall n by default when including both sexes in in vivo studies: evidence from statistical simulations"

Supporting material for manuscript "Scientists should not increase the overall n by default when including both sexes in in vivo studies: evidence from statistical simulations" Includes underlying data and scripts to enable reproducible analysis.

FIG. 5 in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 5. — Fragilaria pectinalis (O.F.Müll.) Lyngb. Images taken from the River Aa, province of Antwerp, Belgium: A-S, LM views of more or less normally formed valves of the population arranged in decreasing length; T, U, LM views of two frustules connected in girdle view; V-AJ, LM views of several deformed valves of the population arranged in decreasing length; AK, SEM external view of a frustule in oblique view showing the girdle bands and the mantle plaques; AL, SEM external view of an entire valve; AM, SEM external detail of a valve apex with the rimoportula; AN, SEM external detail of a valve apex lacking the rimoportula. Note the spines and small granules at the valve margin; AO, SEM internal view of an entire valve showing the slightly eccentric position of the rimoportula at one apex. Scale bars: A-AL, AN, 10 µm; AM, AO, 1 µm.

FIG. 2 in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 2. — Synedra vaucheriae var. deformis Grunow. Images taken from the type material (Grunow sample 907): A, LM view of four frustules connected in girdle view; B-Q, LM views of the population arranged in decreasing length; R, original drawing and notes made by Grunow (Naturhistorisches Museum Wien, Austria); S, SEM external detail of the apex of a frustule girdle view showing the girdle bands and the apical pore fields; T, SEM external view of an entire valve; U, SEM external detail of a valve apex with the rimoportula; V, SEM internal view of an entire valve showing the slightly eccentric position of the rimoportula at one apex. Scale bars: A-Q, T, V, 10 µm; S, U, 1 µm.

FIG. 1. — Synedra deformis W in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 1. — Synedra deformis W.Sm. Images taken from the type material (P. Mill Pond, Lewes, Sussex, United Kingdom): A, LM view of two frustules connected in girdle view; B-R, LM views of the population arranged in decreasing length; S, SEM external view of a frustule in girdle view; T, SEM external view of an entire valve. The arrows indicate interruptions in the areola series of the striae; U, V, SEM external view of two smaller valves. The arrows indicate the rimoportula; W, SEM external detail of the valve apex showing the rimoportula (arrow) and the apical pore field; X, SEM internal view of an entire valve showing the central position of the rimoportula at one apex; Y, SEM internal detail showing the eccentric position of the rimoportula. Scale bars: A-V, X, 10 µm; W, Y, 1 µm.

FIG. 4 in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 4. — Fragilaria joachimii Kahlert. Images taken from the type material (Strain TCC887 isolated from BrostrÖmmen near Norrtälje, Sweden): A-R, LM views of the population arranged in decreasing length; S, T, LM views of several frustules connected to form long chains in girdle view; U, SEM external view of several frustules in girdle view connected via mucus (indicated by the arrows); V, SEM external view of a frustule in girdle view showing the girdle bands and the mantle plaques. Note the small spines and granules near the valve face/mantle junction; W, SEM external view of an entire valve with scattered small spines on the margin; X, SEM external detail of a valve apex with the rimoportula and the apical pore field. Scale bars: A-W, 10 µm; X, 1 µm.

FIG. 6 in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 6. — Fragilaria vaucheriae (Kütz.) J.B.Petersen. Images taken from the type population (Quellen bei Leisling, Weissenfels, Germany): A, B, LM views of two frustules connected in girdle view; C-P, LM views of the population arranged in decreasing length. Scale bar: 10 µm.

FIG. 3 in Analysis of the type material of Synedra deformis W.Sm. and Synedra vaucheriae var. deformis Grunow (Fragilariaceae, Bacillariophyta)

FIG. 3. — Fragilaria candidagilae Almeida, C.Delgado, Novais & S.Blanco. Images taken from the type material (Ribeira do Botão, Mondego River Basin, Coimbra Portugal): A, LM view of two frustules connected in girdle view; B-R, LM views of the population arranged in decreasing length; S, SEM external view of a frustule in girdle view showing the girdle bands, the apical pore fields and the mantle plaques; T, SEM external view of an entire valve; U, SEM external view of an entire small valve; V, SEM external detail of a valve apex with the rimoportula and the apical pore field; W, SEM internal view of an entire valve showing the slightly eccentric position of the rimoportula at one apex. Scale bars: A-U, W, 10 µm; V, 1 µm.

Indoor UWB CIR Data Set for Material Prediction

ABOUT This data set contains spatially distributed CIR of multipath components in indoor environment acquired with ultra wideband (UWB) radio technology in microwave frequency range. The data is labeled with the materials of the surfaces bounding the space (floor, ceiling, walls). The data was collected for training and evaluating machine learning models for CIR-based indoor material prediction, but it may be also used for other studies based on indoor radio propagation data.    AUTHORS Teodora Kocevska, Andrej Hrovat, Tomaž Javornik Department of Communication Systems Jožef Stefan Institute, SI-1000 Ljubljana, Slovenia   DATA COLLECTION The synthetic data is obtained using <a href="https://www.remcom.com/wireless-insite-em-propagation-software">Remcom Wireless InSite</a> v.3.3.3. The CIR is estimated in 16,875 rooms in total.  These rooms belong to 5,625 distinct room types and each room type is considered in three sizes.  The number of distinct room types comes from the materials used for the floor, ceiling, and walls, having nine floor-ceiling material combinations and 625 wall-material combinations.   <table align="center"> <caption>Room Sizes</caption> <thead> <tr> <th scope="col"> ROOM SIZE</th> <th scope="col">FLOOR/CEILING DIMENSIONS </th> <th scope="col">WALL DIMENSIONS</th> </tr> </thead> <tbody> <tr> <td>S</td> <td>3 m x 3 m</td> <td>3 m x 3 m  </td> </tr> <tr> <td>M</td> <td>5 m x 5 m</td> <td>5 m x 3 m </td> </tr> <tr> <td>L</td> <td>7 m x 7 m </td> <td>7 m x 3 m </td> </tr> </tbody> </table>   MATERIALS <ul> <li> floor: concrete, wood, floorboard</li> <li> ceiling: concrete, plaster, wood</li> <li> walls: brick, concrete, glass, plaster, wood</li> </ul>   <table align="center"> <caption>Electrical properties of materials*</caption> <thead> <tr> <th scope="col">MATERIAL</th> <th scope="col">RELATIVE PERMITTIVITY</th> <th scope="col">CONDUCTIVITY</th> </tr> </thead> <tbody> <tr> <td>brick</td> <td>3.75</td> <td>0.038</td> </tr> <tr> <td>concrete</td> <td>5.31</td> <td>0.120</td> </tr> <tr> <td>glass</td> <td>6.27</td> <td>0.029</td> </tr> <tr> <td>plaster</td> <td>2.94</td> <td>0.036</td> </tr> <tr> <td>wood</td> <td>1.99</td> <td>0.026</td> </tr> <tr> <td>floorboard</td> <td>3.66</td> <td>0.039</td> </tr> </tbody> </table>   COMMUNICATION SYSTEM CONFIGURATION Ultra wideband (UWB) radio technology is considered. The parameters of the communication system are set according to 802.15.4-2011** standard. The configuration of system parameters is summarized as follows: <table align="center"> <caption>System configuration</caption> <thead> <tr> <th scope="col"> PARAMETER</th> <th scope="col">CONFIGURATION</th> </tr> </thead> <tbody> <tr> <td>frequency</td> <td>3494.4 MHz</td> </tr> <tr> <td>bandwidth</td> <td>466.2 MHz</td> </tr> <tr> <td>Tx/Rx height</td> <td>1.5 m</td> </tr> <tr> <td>antenna type</td> <td>omni</td> </tr> <tr> <td>polarization</td> <td>vertical</td> </tr> </tbody> </table>   RADIO NODE POSITIONS The data is collected using three acquisition layouts as follows: 1. Layout 1    Tx in the center of the room and Rx moved over uniform grid covering the room. 2. Layout 2 Tx in eight positions following circular pattern around the center of the room and Rx moved over uniform grid covering the room.  The distance from the center of the room to the circumference of the circle is 0.5 m, and the spacing between the radio nodes is pi/4 rad. 3. Layout 3 Tx in four positions near the corners of the room (0.375 m from the walls) and Rx moved over uniform grid covering the room. The corners of the grid are 0.25 m apart from the walls and the distance between the nodes is also 0.25 m. Since the grid size is defined relatively to the room size, the total number of grid node positions is different in rooms with different sizes. The total number of grid node positions is 121, 361, and 729 in S, M, and L rooms, respectively.   DATA ORGANIZATION Data is saved in .csv files. Each file starts with a header line specifying the column names.  The column names included in the .csv files are: - Column 0: layout {center, circle, corners} - Column 1: tx_point_id {1} for Layout 1, {1-8} for Layout 2, and {1-4} for Layout 3 - Column 2: rx_point_id {1-121} in S-rooms, {1-361} in M-rooms, and {1-729} in L-rooms - Column 3: 1_phase_deg NUMERIC - Column 4: 1_toa_ns NUMERIC - Column 5: 1_power_dbm NUMERIC - Column 6: 1_power_nw NUMERIC - Column 7: 2_phase_deg NUMERIC - Column 8: 2_toa_ns NUMERIC - Column 9: 2_power_dbm NUMERIC - Column 10: 2_power_nw NUMERIC - Column 11: 3_phase_deg NUMERIC - Column 12: 3_toa_ns NUMERIC - Column 13: 3_power_dbm NUMERIC - Column 14: 3_power_nw NUMERIC - Column 15: 4_phase_deg NUMERIC - Column 16: 4_toa_ns NUMERIC - Column 17: 4_power_dbm NUMERIC - Column 18: 4_power_nw NUMERIC - Column 19: 5_phase_deg NUMERIC - Column 20: 5_toa_ns NUMERIC - Column 21: 5_power_dbm NUMERIC - Column 22: 5_power_nw NUMERIC - Column 23: 6_phase_deg NUMERIC - Column 24: 6_toa_ns NUMERIC - Column 25: 6_power_dbm NUMERIC - Column 26: 6_power_nw NUMERIC - Column 27: 7_phase_deg NUMERIC - Column 28: 7_toa_ns NUMERIC - Column 29: 7_power_dbm NUMERIC - Column 30: 7_power_nw NUMERIC - Column 31: 8_phase_deg NUMERIC - Column 32: 8_toa_ns NUMERIC - Column 33: 8_power_dbm NUMERIC - Column 34: 8_power_nw NUMERIC - Column 35: 9_phase_deg NUMERIC - Column 36: 9_toa_ns NUMERIC - Column 37: 9_power_dbm NUMERIC - Column 38: 9_power_nw NUMERIC - Column 39: 10_phase_deg NUMERIC - Column 40: 10_toa_ns NUMERIC - Column 41: 10_power_dbm NUMERIC - Column 42: 10_power_nw NUMERIC - Column 43: 11_phase_deg NUMERIC - Column 44: 11_toa_ns NUMERIC - Column 45: 11_power_dbm NUMERIC - Column 46: 11_power_nw NUMERIC - Column 47: 12_phase_deg NUMERIC - Column 48: 12_toa_ns NUMERIC - Column 49: 12_power_dbm NUMERIC - Column 50: 12_power_nw NUMERIC - Column 51: 13_phase_deg NUMERIC - Column 52: 13_toa_ns NUMERIC - Column 53: 13_power_dbm NUMERIC - Column 54: 13_power_nw NUMERIC - Column 55: 14_phase_deg NUMERIC - Column 56: 14_toa_ns NUMERIC - Column 57: 14_power_dbm NUMERIC - Column 58: 14_power_nw NUMERIC - Column 59: 15_phase_deg NUMERIC - Column 60: 15_toa_ns NUMERIC - Column 61: 15_power_dbm NUMERIC - Column 62: 15_power_nw NUMERIC - Column 63: room_size_surf_m2 {9} for S-rooms, {25} for M-rooms, and {49} for L-rooms - Column 64: room_size_name {S} for S-rooms, {M} for M-rooms, and {L} for L-rooms - Column 65: room_shape {square} - Column 66: floor_mat {concrete, wood, floorboard} - Column 67: ceiling_mat {concrete, plaster, wood} - Column 68: wall1_mat {brick, concrete, glass, plaster, wood} - Column 69: wall2_mat {brick, concrete, glass, plaster, wood} - Column 70: wall3_mat {brick, concrete, glass, plaster, wood} - Column 71: wall4_mat {brick, concrete, glass, plaster, wood} Each row corresponds to separate radio link defined with the Tx and Rx nodes.  It includes information about      (i)  the CIR-acquisition layout (position of the Txs and Rxs),      (ii)  the Tx and Rx,      (iii)  the CIR of 15 strongest multipath components,      (iv)  the room geometry, and      (v)  the materials of the surfaces bounding the space.  - Column 0 specifies the layout.      The following maping is used:          Layout 1 -> center,          Layout 2 -> circle, and          Layout 3 -> corners. - Column 1 specifies the Tx identifier. - Column 2 specifies the Rx identifier. - Columns 3-62 are the input attributes.      The input attributest represent the phase (in deg), ToA (in ns), received power (in dBm), and received power (in nW) for 15 strongest multipath components.      The column naming is X_Y_Z, where X is the multipath component identifier (1 to 15), Y is the propagation characteristic (phase, toa, or power), and Z is the unit (deg, ns, dbm, or nw). - Column 63 specifies the surface of the room in square meters. - Column 64 specifies the room size category (S, M, or L). - Column 65 specifies the room-base shape. - Columns 66-71 are the target attributes specifying the material of the floor, ceiling, wall 1, wall 2, wall 3, and wall 4, respectively.    FOLDER STRUCTURE The folder indoor_CIR_data contains:     - one subfolder named CIR_data         It contains three .csv files with CIR data named by the size of the rooms where the data is acquired.         For naming the .csv files the following mapping is considered:             - Small.csv  -> S-rooms             - Medium.csv -> M-rooms             - Large.csv  -> L-rooms     - one subfolder named CIR_acquisition_details         It contains .png file with schematic representation of the radio node positions considered for obtaining the data.     - README.txt file The folder structure is:     - CIR_data         - Small.csv         - Medium.csv         - Large.csv     - CIR_acquisition_details         - radio_node_positions.png     - README.txt   REFERENCES * R. sector of International Telecommunication Union (ITU-R), “Effects of building materials and structures on radio wave propagation above about 100 MHz,” International Telecommunication Union, ITU-R Recommendation P.2040-2, 2021. ** IEEE, “Standard for local and metropolitan area networks–Part 15.4: Low-rate wireless personal area networks (LR-WPANs),” IEEE, Standard IEEE 802.15.4-2011, 2011.