What's New in Version 5.1
Math library
- Support for quadruple precision numbers with up to 34 digits,
including our full vector and matrix library.
Real and complex numbers are supported.
- Smoothing cubic splines, monotonicity preserving splines.
- Nonlinear curve fitting for arbitrary domains, including surface fitting.
- Improved Quadratic Programming solver.
Vector and Matrix Library
- GPU computing: offload computations to the GPU.
- Data is kept on the GPU as long as possible for optimal performance.
- Row and column labels for vectors and matrices.
- Automatic alignment on labels.
- Grouping and aggregation on vectors and matrices.
- Optimized elementary functions on vectors and matrices.
- Reshaping operations on vectors and matrices.
- Support for F# 3.1 indexing and slicing.
Data Frame Library
- Data frame class supporting advanced data manipulation.
- Simple and multi-level indexes.
- Pivot tables, stacking, unstacking of data frames.
- Joins and merges, join to nearest.
- Advanced grouping and aggregation.
- Supports LINQ queries.
- Integrated with statistical models.
Statistics Library
- GARCH models (Generalized AutoRegressive Conditional Heteroskedasticity).
- Regularized linear regression models: Ridge regression, LASSO and elastic net.
- Cumulative multivariate normal distribution.
- Hyperbolic and Inverse Gaussian distribution.
- Generic truncated distributions.
- Post hoc tests for one-way ANOVA: Tukey, Tukey-Kramer, Scheffé, Fisher-Hayter
- Ljung-Box test for autoregression.