Dew.Stats.Linux is the Linux-native accelerated edition of the Dew.Stats statistical computing library. It provides a comprehensive suite of tools for probability distributions, hypothesis testing, regression, multivariate analysis, experimental design, and time-series modeling, powered by the high-performance vectorized numerical backend in Dew.Math.Linux. This edition is designed for Linux-based compute servers, HPC pipelines, analytics microservices, research clusters, data acquisition systems, and real-time industrial environments. Statistical Capabilities: - Probability distributions (PDF, CDF, inverse CDF) for 36+ discrete and continuous models - Random number generators and parameter estimation - Descriptive statistics, histograms, ogives, quantiles, outlier analysis Hypothesis Testing and Inference: - Parametric tests (t, Z, F, Chi-Squared, Bartlett, Hotelling T²) - Non-parametric tests (Wilcoxon, Sign, Mann–Whitney, Anderson–Darling, Shapiro–Wilk, KS) - Confidence intervals, residual diagnostics, model fitness evaluation Regression and Statistical Modeling: - Linear, multiple linear, logistic, Poisson, ridge and nonlinear regression - ANOVA and ANCOVA - Principal Component Regression and regularization workflows Multivariate and Structural Analysis: - PCA (correlation/covariance) with eigen decomposition - PCA residuals, factor rotation, Bartlett tests, item analysis - Classical Multidimensional Scaling and dimensionality interpretation Time Series Modeling and Forecasting: - ACF and PACF analysis - ARMA, ARIMA and ARAR models - Exponential smoothing (single/double/triple) - Box-Ljung significance testing and forecasting evaluation High-Level Statistical Workflow Components: - TMtxANOVA, TMtxMulLinReg, TMtxNonLinReg, TMtxPCA, TMtxHypothesisTest, TMtxBinaryTest, TMtxMDScaling Platform Characteristics: - Uses **Dew.Math.Linux** for native BLAS/LAPACK acceleration with AVX2/AVX512 dispatch - Highly scalable under multi-threaded workloads - No Windows or WinForms dependencies - Headless execution suitable for batch, service, and compute-node environments Dew.Stats.Linux provides the full analytical capabilities of Dew.Stats, optimized specifically for Linux-based CPU compute environments.
Get Started
$ dotnet add package Dew.Stats.LinuxReadme
Dew.Stats.Linux
Advanced statistical and data analysis library built on top of the
Dew Math Linux Library.
The library provides:
- Probability distributions: PDF, CDF, inverse CDF for 36+ distributions, random number generators, and parameter estimation.
- Descriptive statistics: histograms, ogives, cumulative sums, statistical moments, percentiles, range, IQR, mean, median, mode, ranking tools, and more.
- Multivariate analysis: PCA (covariance/correlation matrix), PCA residuals, orthogonal rotation of Z-scores, Bartlett test for dimensionality, Z-Scores, classical multidimensional scaling, Hotelling T² test, M-Box test, item analysis.
- Design of experiments: full factorial designs, Latin Hypercube sampling.
- Hypothesis testing: Sign test, Wilcoxon Signed-Rank, t-tests (one-sample and two-sample), Z-test, Chi-Squared tests, F-test, Shapiro-Wilk, Shapiro-Francia, Anderson-Darling, Kolmogorov-Smirnov, Mann-Whitney U, Lilliefors, Jarque-Bera, and others.
- Regression models: linear (weighted/unweighted), multiple regression, logistic regression, ridge regression, Poisson regression, general non-linear regression (BFGS, Marquardt, Conjugate Gradient, Simplex), one-way and two-way ANOVA, principal component regression.
- Time series analysis: ACF/PACF, exponential smoothing (single/double/triple), ARMA/ARIMA (simulation, estimation, forecasting — Yule-Walker, Burg, Innovations, MLE), ARAR modeling, moving averaged filters, Box-Ljung statistics.
- Ready-to-use components:
TMtxANOVA– ANOVA routinesTMtxMulLinReg– multiple linear regressionTMtxNonLinReg– nonlinear regressionTMtxPCA– principal component analysisTMtxHypothesisTest– hypothesis testingTMtxBinaryTest– binary diagnostic performance testingTMtxMDScaling– multidimensional scaling
Linux builds do not include TeeChart support.
Examples
- GitHub organization: https://github.com/Dew-Research
- WinForms / .NET Samples: https://github.com/Dew-Research/DewLabStudio-NETCore-Samples
Documentation
- API Reference: https://www.dewresearch.com/Help/VSNET/StatsMaster/contents.html
- Product Page: https://www.dewresearch.com/products/stats-master/stats-master-for-visual-studio-net
- NuGet profile: https://www.nuget.org/profiles/DewResearch/
- License (PDF): https://www.dewresearch.com/downloads/LicenseNET.pdf
License
Commercial license. See DewStatsLicense.txt included in this package.
All Dew.* NuGet packages on nuget.org are trial versions and work from inside Visual Studio IDE only.