This is a dataset prepared and intended as a data source for development of stress analysis methods based on machine learning. The dataset is based on finite element (FEM/FEA) stress analyses of generated mechanical structures using PyCalculix Python API. The dataset contains more than 270,794 pairs of stress analyses images (von Mises stress) of randomly generated 2D structures with predefined thickness and material properties. All the structures are fixed at their bottom edges and loaded with gravity force only. See PREVIEW directory with some examples.
Contributors:del Brio Dolores, Reeb Pablo Daniel, Fernández Dario
The dataset is composed of 60 pear tree images. They were taken during full bloom in a pear orchard located at National Institute of Agricultural Technology (INTA) Experimental Station, General Roca, Argentina (39° 1’ 40’’ S; 67° 44’ 34’’ W).The orchard was established in 2003 with ‘Williams’ cultivar pear trees grafted on seedling rootstock .Trees were planted in a total area of 1.8 ha at a distance of 2m between trees by 4m between rows and were trained as espalier. A few days before full bloom, images were obtained from 30 trees, under two conditions: i) natural daylight between 10 am and 13 pm (PE_FL_DA_2018), ii) at night with the artificial flash light of the camera (PE_FL_NI_2018). A black curtain was unfolded behind the trees when images were obtained under daylight conditions in order to avoid interference from neighboring trees. All images were taken with an RGB digital camera (14.1 MP) at approximately 3.0 m from the tree in a 90° angle to the row. An object of known dimensions (a 15x15 cm square) was placed in each tree as a scale reference. Simultaneously, all the flower clusters on each tree were manually counted. Images taken by using different proximal sensors can be used to estimate the number of flowers or fruits in trees. The accuracy of those methods has been studied and tested in different fruit species by many researchers with encouraging results. A similar dataset has been published by the same contributors in Mendeley Data Repository for apple trees.
This database contains facial images of volunteers in frontal and random poses. Each facial image collection has a visible light image, an infrared image and a depth image.
The images in this database were collected by a single person during the period of October 10 to 27 of the year 2017. 8 classes in the computer science course at UNIVALI University were invited to participate in this paper, where 64 accepted to participate in this work.
The database contains:
267 facial image samples
267 facial images of the visible light spectrum
267 facial images of the infrared spectrum
267 depth images (3D)
Seven sets of c Gaussian shaped clustered datasets. For each dataset, n points with p dimensions were generated from a mixture of c Gaussian distributions (clusters). The means of each cluster were randomly generated with numbers between 0 and 10 and they were structured in a c x p matrix. The standard deviations of each cluster are represented by a p x p covariance matrix generated by a normally distributed random numbers. After all, from the matrix of means c x p and the c covariance matrices p x p, the mvrnorm function of the MASS library of the R software was used to produce the samples that composes the Gaussian mixture dataset.
The properties of the seven Gaussian mixture datasets:
Dataset | p | n | c
Gaussian.k8 | 657 | 81 | 8
Gaussian.k2 | 4,232 | 181 | 2
Gaussian.k7 | 4,514 | 128 | 7
Gaussian.k9 | 5,041 | 108 | 9
Gaussian.k6 | 5,176 | 143 | 6
Gaussian.k5 | 6,203 | 130 | 5
Gaussian.k4 | 6,615 | 168 | 4
The cluster label of each object is in the last column of the dataset.
Each column is separated by a comma and there are not columns and row names.
Side reactions involving surface reduction play a critical role in the failure of LiNi0.8Co0.15Al0.05O2 to reach its theoretical capacity as a cathode material for Li-ion batteries. While macroscopic consequences are known, the underlying nanoscopic mechanisms are not fully elucidated. By coupling X-ray spectroscopy with several X-ray microscopy modalities, we have spatially resolved the extent of Ni oxidation at several states of charge and uncovered heterogeneity that is hidden when considering ensemble measurements alone. The use of morphologically controlled particles enabled high-resolution imaging of these materials, uncovering gradients of Ni oxidation states within individual primary particles. At high states of charge, these gradients revealed regions of possible oxygen deficiency extending deeper into the particle than previously observed. Surface-sensitive X-ray coupled scanning tunneling microscopy allows oxidation states to be measured at the material’s surface, showing predominantly Ni(II) in the first atomic layer, and mixtures of Ni(II) with Ni(III) / Ni(IV) already appearing 1.5nm into the particle. These results reveal the subtle interplay between irreversible surface transformations and the bulk reactions that ultimately define function, which will refine strategies of surface passivation that are key to overcoming current performance limitations.
Contributors:Shi Yuling, Yi Xuemei, Zhang Yi, Yu Ning, Yu Qian, Le Qianlian, Chen Zeyu, Ding Yangfeng, Yuling Shi
Supplemental Figure 1. Treatment regimen.
Supplemental Table 1. Inclusion and exclusion criteria
Supplemental Table 2. Protocol and adjustments
Supplemental Table 3. Patient characteristics at baseline
Supplemental Table 4. Primary, ranked secondary, and additional endpoints
Supplemental Table 5. Comparison of PASI 75 and PASI 90 response rate between patients with or without tendency for remission in summer
Contributors:Henrique dos Santos Elineuma , Barni Paulo Eduardo , Rodrigues Silva Williamar, Imbrozio Barbosa Reinaldo
The dataset provides information on biomass / carbon estimates for the ecotone forests dispersed on the eastern region of the Maracá Ecological Station; a Brazilian protected area formed by Maracá Island and islets (hereafter “Maracá”) located on the Rio Uraricoera, State of Roraima, northern Brazilian Amazon. The carbon in above-ground total biomass (AGB) was estimated based on data from the 4th “tree+palm” census (https://data.mendeley.com/datasets/8cdwkhcsy7/2) carried out in 129 permanent plots (50 mx 10 m; 6.45 ha) installed on the six East-West trails in the PPBio research grid installed in Maracá (https://ppbio.inpa.gov.br/sitios/maraca). A general allometric model (Chave et al., 2014) was adopted for estimate AGB for each tree in the database. The palms biomass was performed at the gender level using the allometric models of Goodman et al. (2013). The individual carbon was calculated multiplying the biomass of each tree/palm by a factor of 0.5 (considering 50% of C). The individuals carbon values were used to compose the total carbon stock in each plot taking into account three arboreal classes: (i) P. gracilipes (Leguminosae tree species with higher abundance in the region), (ii) others (all other tree species) and (iii) palms (a set of five arborescent species). The dataset consists of three files: (i) study_site - a figure indicating the geographical location of the study area; (ii) location_plot - identification of the sampling plots (unique code) and their geographical location in (i) lat / long - WGS84 and (ii) UTM / Zone 20 N, WGS 84; (iii) biomass_carbon - biomass and carbon stocks of P. gracilipes, others and palm trees (Mg / 500m² and % of each arboreal class). The current dataset was supported by the projects (i) SavFloRR - Ecologia e manejo dos recursos naturais de ecossistemas de savanas e florestas de Roraima (PPI INPA 015/122), and (ii) Crescimento e mortalidade de árvores em florestas ecotonais de Roraima: efeito das condicionantes ambientais e da variabilidade climática (Proc. CNPq n. 403591 / 2016-3). The Coordination for the Improvement of Higher Education Personnel (CAPES) supported E.H.S. The PELD Program provide a scholarship to W.R.S. (CNPq / CAPES / FAPs / BC-Fundo Newton; Proc. N. 441575 / 2016-1). The National Council Scientific and Technological Development (CNPq) provided a grant to R.I. Barbosa (CNPq 304204 / 2015-3). The Chico Mendes Institute for Biodiversity Conservation (ICMBio) authorized the study (SISBIO nº 52071).
Contributors:Teong Kai Lun, H.G. Tan Rodney, Y.H. Hoo Landon
Measurement datasets are attached in Microsoft workbook format, entitled as:
1)Mono70W 999 Irradiance
2)Poly70W 1000 Irradiance
which are the experimentally measured I-V curves for both mono- and poly- crystalline silicon PV modules. The datasets are obtained by connecting the PV Analyzer towards the targetted PV module, placing the PV module directly under the sun, in which such experiment is conducted in the afternoon. The generated reports from the PV Analyzer's software are filtered, and two of the datasets are selected to present in this paper.
Other than the measured specification of the PV module, the datasets also consist three columns of data, which are the voltage, current and power. By plotting these data points accordingly towards x- and y- axis, the experimentally measured I-V and/or P-V curve is obtained.
Total of 4 main programs developed through MATLAB platform are attached, entitled as:
which 1) and 2) are the I-V curve fitting parameterization algorithm with double error criteria respectively for proposed and traditional mathematical model; while 3) and 4) are the same algorithm with single error criterion. The rest of the .m files are coded to complement the features provided in the main programs.
The program entitled "WinOnTop" is obtained from https://www.mathworks.com/matlabcentral/fileexchange/42252-winontop, thus cited as:
Igor (2020). WinOnTop (https://www.github.com/i3v/WinOnTop), GitHub. Retrieved June 6, 2020.
By executing the programs in MATLAB, press the "Load File" button to select the I-V curve dataset. The user can adjust the Rs, A, and Rsh parameters by pressing the interface button, or pressing the "Auto" button to execute the Cyclical Rs-A-Rsh parameterization algorithm that programmed by the author targeting the objective of minimizing the MAEP error between measured and modeled I-V curve. While the three parameters are adjusting, the information will be updated accordingly. When the "Auto" curve-fitting is completed, a report can be generated showing the successive MAEP iteration throughout the curve fitting if the user presses the "OK" button on the generated window.
A folder entitled "Result" is attached, including four .m files and eight .fig files, which all of these are presented in the result section of the paper.
Four main programs entitled:
are attached to compare the obtained results. Comparisons are made between the traditional and proposed mathematical models. 1) and 2) compare the result for mono- crystalline silicon PV module respectively for double error criteria and single error criterion; while 3) and 4) compare the same error criterion for poly- crystalline silicon PV module.
The high-speed camera images of HY-30-90-A used in this paper, including the reference image, region of interest and other 177 images in sequence inTIF format. The image numbered 68545 is the first frame with macro cracks. The original data collected by the high-speed camera is too large to include, and is available upon request.