Stata - Matrix Programming : Mata

New in Stata 11

Environment

  • Interactive—type matrix expressions and see results
  • Development—full development environment for developing programs and commands
  • Automatically compiled code

General features and matrix operators

  • Support for real and complex values
  • Joining by rows or columns
  • Addition, subtraction, multiplication, scalar division
  • Elementwise arithmetic
  • Transposition and reshaping
  • Kronecker and Hadamard products
  • Inner and outer products

Optimizer (Updated)

  • Code the function, function and gradient, or function, gradient, and Hessian
  • Newton–Raphson, BFGS, DFP, BHHH, Nelder–Mead, and Gauss–Newton techniques
  • Versions for general problems and statistical problems like maximum likelihood New
  • Debugger
  • Covariance matrix based on Hessian, outer product of gradients, or robust/sandwich

Mathematical and matrix functions

  • LAPACK numerical analysis routines (Updated)
  • Symmetric, nonsymmetric, and generalized inverses
  • Cholesky, LU, QR, and SVD solvers
  • Transcendental and trigonometric functions
  • Gamma and factorial functions
  • Density and distribution functions
  • Polynomial evaluation, arithmetic, and calculus
  • Hilbert, Toeplitz, and Vandermonde matrices
  • Duplication, commutation, and elimination matrices New
  • Random-number generation
  • Fast Fourier analysis
  • Cubic splines
  • Means, sums, minimums, and maximums of matrices
  • Numerical first and second derivatives New

Object-oriented programming New

  • Inheritance
  • Virtual functions
  • Encapsulation of data structures and programs

Decompositions

  • Eigenvalues and eigenvectors (Updated)
  • Cholesky
  • LU and LUD
  • QR
  • Schur New
  • Hessenberg New
  • Singular value
  • Permutation matrices

Data accumulation

  • Cross products
  • Cross products of deviations from means
  • Weighted and GLS-type cross products
  • Support for quad-precision cross products

Stata interface

  • Convert Stata datasets to matrices and vice versa
  • Virtual matrix views onto dataset
  • Access and set estimation-class and other macros
  • Manipulate dataset value labels

Other features

  • String scalars and matrices
  • File I/O, including buffered I/O
  • C-like syntax
  • Fast execution










Back to Capabilities Home

Back to Stata 11 Overview


Back to Stata homepage
Back to Timberlake Consultants

©Timberlake Consultants Limited
Last revised:07/08/2009