Small dense linear algebra for the regression modules: row-major Float64 matrices, Householder QR
least squares with rank detection, Cholesky, general inverse, and a one-sided Jacobi SVD.
Sizes are n × p with p small (tens of columns), so O(np²) algorithms are used throughout.
Small dense linear algebra for the regression modules: row-major Float64 matrices, Householder QR least squares with rank detection, Cholesky, general inverse, and a one-sided Jacobi SVD. Sizes are n × p with p small (tens of columns), so O(np²) algorithms are used throughout.