Find research datasets worth reusing

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 4. Statistical Metrics distribution of model on RIO as response of Terasort application

Figure 4 shows the statistical metrics distribution of regression model on read rate as the response of Terasort application. The filled triangle point-up indicates the minimum stable sampling time for statistical metrics. The top-half of figure 4 shows the residual standard error (RSE) distribution as training data size increase. The remaining half is for the distribution of R2.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 3. Estimate coefficients distribution of model on read rate as response of Terasort application

Figure 3 shows estimate coefficients distribution of model on read rate as the response of Terasort application. The filled triangle point-up indicates the position of the minimum stable sampling time for the corresponding estimate coefficients as well as the number above it shows the exact position value. The dashed line represents the average estimated coefficient of the regression model.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 1. ACF and PACF plot of TeraSort application

In Figure 1, the autocorrelation plots of all resource usage parameters reveal non- randomness because of the corresponding autocorrelations, denoted by the circle, violate the dashed lines (95% confidence boundary) and are statistically significant for lags up to 100.  The filled triangle point-up in partial autocorrelation plots marks the largest partial autocorrelation of usage parameters as well as the corresponding lag number.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 6. RSE and R2 of regression models of MapReduce applications

Figure 6 shows the fit quality of regression models. It is following.The left panel and the right panel of figure 6 show the residual standard error (RSE) distribution and R2 distribution of each application. The good fit quality corresponds to a taller R2 bar and a shorter RSE bar. The R2 almost 1 and small RSE show the best fit quality of the regression models on memory usage as the response. The overall higher RSE and lower R2 of regression models on CPU as the response show the worse quality of fitting goodness. The regression models on read rate as the response also show a moderate fitting quality. For the regression models on write rate as the response, Terasort application exhibits the best quality and Teragen application as well. Others show the worse fitting quality. The results show that the regression models on intensive usage parameters as response exhibit the good fitting quality.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 2. ACF plot of Residuals of Regression Model of Terasort

, are uncorrelated. If the error term is uncorrelated, it proves that there exists strong randomness in residuals of the model and provides the evidence for the unbiased estimate for the true standard error. The autocorrelation plot is used to check this assumption. Figure 2 shows the autocorrelation plot of residuals of regression models of TeraSort application. In Figure 2, the horizontal axis represents lag time and the vertical axis indicates the autocorrelation between residual at time t and residual at other lag time. At lag 0, autocorrelation is always equal to 1 and represents time series itself. Most of the autocorrelation at other lag time fall into the 95% confidence interval, only few of them violate the dashed line. Such a shape of ACF plot proves that residuals are uncorrelated and respects to the independent assumption of the linear regression model residuals.

opencc-by-4.0May 2018View details →

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 8. Minimum sample time of statistical metrics of MapReduce applications

Figure 8 presents the minimum sampling time distribution of statistic metrics which ensures the stable modeling. Overall, the minimum sampling time of statistic metrics is smaller than sampling time of estimated coefficients. For different applications, a time-consuming application like Terasort needs the largest sampling time to tend to be stable. The Pi application shows the smallest minimum sampling time to reach stability.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Figure 7. Minimum sampling time of MapReduce applications on various response variable

Pi application needs the smallest sampling time. The remaining applications need the similar minimum sampling time. Overall, the stable regression models on memory usage as response show the least need for sampling time. The results show that various applications have different minimum sampling time to get stable. The application which performs more read/write operations shows larger sampling time need.

Stable Modeling on Resource Usage Parameters of MapReduce Application-Department of Networked Systems and Services, Budapest University of Technology and Economics, Budapest, Hungary

In Figure 5, the positive dependency of different strength between each resource usage parameter and the corresponding previous usage parameter is exhibited for all MapReduce applications. It indicates that all current resource usage parameters are positively dependent on the previous values to some extent degree. Except for these common dependencies, there exist some special dependencies for different applications. On the top-left panel of Figure 5, CPU usage of Pi application shows the strongest positive dependency to lagged CPU usage, the Teragen application had the weakest positive dependency, and others exhibit the moderate positive dependency.

opengpl-2.0Jun 2018View details →

Companion for "Design, Implementation and Performance Analysis of a CFD task-based Application for Heterogeneous CPU/GPU Resources"

This is the companion data for the VECPAR2018 submission paper entitled: Design, Implementation and Performance Analysis of a CFD task-based Application for Heterogeneous CPU/GPU Resources by Lucas Leandro Nesi, Lucas Mello Schnorr, and Philippe Olivier Alexandre Navaux. All the data, source code, and images generation scripts used in the paper are present here.

opencc-by-4.0Feb 2018View details →

Avoiding sedentary behaviors requires more cortical resources than avoiding physical activity

Dataset related to the paper entitled "Avoiding sedentary behaviors requires more cortical resources than avoiding physical activity".  This dataset includes: 1) A workbook 2) Raw data ("raw_data_eprime_zen.csv") of the behavioral outcomes of the manikin task 3) Self-reported data ("data_self_report_R_subset_zen.csv"). 4) Electroencephalography data ("ERP_by_subject_by_condition_data.zip"). 5) R script for the data management of the behavioral outcomes (i.e., from the raw data to data ready to be analyzed) 6) Images used in the manikin task 7) Eprime script for the manikin task

AGRIS resources for stevia rebaudiana

This dataset includes bibliographic resources from the AGRIS database related to stevia rebaudiana

Agriculture - General: Natural Resources and Environment 2

Agriculture - General: Natural Resources and Environment

Original data comes from a project which takes or took place as part of the DFG priority program “Exploratories for large-scale and long-term functional biodiversity research”. The data is stored together with descriptive metadata, in combination called a dataset, in the project repository (<a href="https://www.bexis.uni-jena.de">https://www.bexis.uni-jena.de</a>). Species information was extracted from that original dataset. The second paragraph is part of the metadata of the original dataset.‘Sammelarten" sind durch agg= Aggregate oder total gekennzeichnet. Sie umfassen jene Taxa, die teilweise bis zur Kleinart, Subspecies oder Varietas bestimmt wurden, teilweise aber auch nur auf Gattungs- (… spec. (indet.)) oder Artebene. Taxonomie nach Wisskirchen,Haeupler (1998)Standardliste für Deutschland. Lebensformen nach Raunkiaer, C. (1910): Statistik der Lebensformen als Grundlage für die biologische Pflanzengeographie.-Beih. Biol. Cbl. 27 II, 170 - 206d. Rote Liste: KORNECK, D., SCHNITTLER, M., VOLLMER, I. (1996): Rote Liste der Farn- und Blütenpflanzen (Pteridophyta et Spermatophyta) Deutschlands. – LUDWIG, G., SCHNITTLER, M. [Hrsg.]: Rote Listen gefährdeter Pflanzen Deutschlands. – Schriftenr. Vegetationskd. 28: 21–187, Bundesamt für Naturschutz, Bonn.’ Socher S (2016). vegetation releves on grassland gridplots 2007-2009. Biodiversity Exploratories. Occurrence dataset <a href="https://doi.org/10.15468/dmjxfo">https://doi.org/10.15468/dmjxfo</a> accessed via GBIF.org

Agriculture - General: Natural Resources and Environment, Plant Production and Protection

opencc-by-4.0Aug 2018View details →

China's transboundary water resources estimation using machine learning approaches

China's transboundary water resources estimated using machine learning models (random forest, gradient boosting, and stacking).

opencc-by-4.0Nov 2018View details →

Back to the edge: relative coordinate system for use-wear analysis [complement to Online Resource 3]

3D micro surface data processed in ConfoMap v7.4.8633 (a derivative of MountainsMap Imaging Topography developed by Digital Surf, Besançon, France). Instructions to download all files at once are given here: <a href="https://doi.org/10.5281/zenodo.4011952">https://doi.org/10.5281/zenodo.4011952</a>

opencc-by-4.0Nov 2017View details →

Secondary metabolites from nectar and pollen: a resource for ecological and evolutionary studies

Floral chemistry mediates plant interactions with herbivores, pathogens, and pollinators. The chemistry of floral nectar and pollen—the primary food rewards for pollinators—can affect both plant reproduction and pollinator health. Although the existence and functional significance of nectar and pollen secondary metabolites has long been known, comprehensive quantitative characterizations of secondary chemistry exist for only a few species. Moreover, little is known about intraspecific variation in nectar and pollen chemical profiles. Because the ecological effects of secondary chemicals are dose-dependent, heterogeneity across genotypes and populations could influence floral trait evolution and pollinator foraging ecology. To better understand within- and across-species heterogeneity in nectar and pollen secondary chemistry, we undertook exhaustive LC-MS and LC-UV-based chemical characterizations of nectar and pollen methanol extracts from 31 cultivated and wild plant species.  Nectar and pollen were collected from farms and natural areas in Massachusetts, Vermont, and California, USA, in 2013 and 2014. For wild species, we aimed to collect 10 samples from each of 3 sites. For agricultural and horticultural species, we aimed for 10 samples from each of 3 cultivars. Our dataset (1535 samples, 102 identified compounds) identifies and quantifies each compound recorded in methanolic extracts, and includes chemical metadata that describe the molecular mass, retention time, and chemical classification of each compound. A reference phylogeny is included for comparative analyses. We found that each species possessed a distinct chemical profile; moreover, within species, few compounds were found in both nectar and pollen. The most common secondary chemical classes were flavonoids, terpenoids, alkaloids and amines, and chlorogenic acids. The most common compounds were quercetin and kaempferol glycosides. Pollens contained high concentrations of hydroxycinnamoyl-spermidine conjugates, mainly triscoumaroyl and trisferuloyl spermidine, found in 71% of species. When present, pollen alkaloids and spermidines had median nonzero concentrations of 23,000 µM (median 52% of recorded micromolar composition). Although secondary chemistry was qualitatively consistent within each species and sample type, we found significant quantitative heterogeneity across cultivars and sites. These data provide a standard reference for future ecological and evolutionary research on nectar and pollen secondary chemistry, including its role in pollinator health and plant reproduction.

opencc-by-4.0Feb 2019View details →

Freshwater resources under success and failure of the Paris climate agreement

This dataset represents the core output of the analysis presented in: Heinke, J., Müller, C., Lannerstad, M., Gerten, D., and Lucht, W.: Freshwater resources under success and failure of the Paris climate agreement, Earth Syst. Dynam., 10, 205-217, 10.5194/esd-10-205-2019, 2019. Please refer to this publication for a comprehensive description of methods and references to the datasets and materials used to produce this data. When using the data, cite it as follows: Heinke, Jens, Müller, Christoph, Lannerstad, Mats, Gerten, Dieter, & Lucht, Wolfgang (2019). Freshwater resources under success and failure of the Paris climate agreement [Data set]. Zenodo. http://doi.org/10.5281/zenodo.2562056. Please also cite the reference article that this dataset belongs to.   Files: <ul> <li>frac_drought_19gcm_8gmt.nc contains the fraction of drought months for 19 GCM patterns and 8 levels of global mean temperature increase</li> <li>frac_drought_ref.nc contains the fraction of drought months for the reference case (equivalent to global mean temperature increase in 2009)</li> <li>mean_annual_discharge_19gcm_8gmt.nc contains mean annual discharge for 19 GCM patterns and 8 levels of global mean temperature increase</li> <li>mean_annual_discharge_ref.nc contains mean annual discharge for the reference case (equivalent to global mean temperature increase in 2009)</li> <li>q10_19gcm_8gmt.nc contains 5-day average peak flow exceeded in 1 of 10 years for 19 GCM patterns and 8 levels of global mean temperature increase</li> <li>q10_ref.nc contains 5-day average peak flow exceeded in 1 of 10 years for the reference case (equivalent to global mean temperature increase in 2009)</li> <li>water_crowding_1950-2010.nc contains estimates of grid-based water crowding for the historic period (1950-2010)</li> <li>water_crowding_2011-2100_5ssp.nc contains estimates of grid-based water crowding for the scenario period (2011-2100) for 5 different SSPs.</li> </ul> All data have a spatial resolution of 0.5° x 0.5° and cover the global land area except Antarctica.

opencc-by-4.0Dec 2018View details →

Data for: Resource landscapes explain contrasting patterns of aggregation and site fidelity by red knots at two wintering sites

This repository contains data for the paper: Oudman et al. 2018. Resource landscapes explain contrasting patterns of aggregation and site fidelity by red knots at two wintering sites. Movement Ecology 6(14) 1-12. https://doi.org/10.1186/s40462-018-0142-4. Please cite the original publication when using this data.