RSpectra: Solvers for Large-Scale Eigenvalue and SVD Problems

R interface to the 'Spectra' library <https://spectralib.org/> for large-scale eigenvalue and SVD problems. It is typically used to compute a few eigenvalues/vectors of an n by n matrix, e.g., the k largest eigenvalues, which is usually more efficient than eigen() if k << n. This package provides the 'eigs()' function that does the similar job as in 'Matlab', 'Octave', 'Python SciPy' and 'Julia'. It also provides the 'svds()' function to calculate the largest k singular values and corresponding singular vectors of a real matrix. The matrix to be computed on can be dense, sparse, or in the form of an operator defined by the user.

Downloads:

Reverse dependencies:

Linking:

Please use the canonical form https://CRAN.R-project.org/package=RSpectra to link to this page.