Provides a linear algebra system used for numerical operations in Tribuo.
There are Dense and Sparse vectors and Matrices, along with a DenseSparseMatrix which is a dense array of sparse row vectors. The dense matrix provides various factorization methods in addition to matrix-vector operations.
It's a single threaded implementation in pure Java. We're looking at ways of improving the speed using new technologies coming in future releases of Java.
ClassDescriptionA dense matrix, backed by a primitive array.The output of a successful Cholesky factorization.The output of a successful eigen decomposition.The output of a successful LU factorization.A matrix which is dense in the first dimension and sparse in the second.A dense vector, backed by a double array.Interface for 2 dimensional
Tensors.Interface for matrix factorizations.A
MatrixTuples.A mutable tuple used to avoid allocation when iterating a matrix.Interface for 1 dimensional
Tensors.A sparse vector.An interface for Tensors, currently Vectors and Matrices.A
VectorTuples.A mutable tuple used to avoid allocation when iterating a vector.