Linear Algebra: Vectors, Matrices, Stats, Model Fitting
Final includes a small dense float64 linear-algebra API for data you already hold as numbers: a vector type, a rank-2 matrix type, stack operators on those values, and ordinary least-squares (OLS) fit / compare / predict.
This is not a growable List. VecNf and Mat2f have a length (and matrix shape) fixed at construction. Use FloatList when you need append, insert, or a general numeric sequence; convert at the boundary when you are ready to compute.
| Page | Contents |
|---|---|
| Vectors and Matrices | VecNf / Mat2f: create, fill, index, iterate |
vec:: and mat:: Operators | First-wave arithmetic, product, solve, row/col, copy |
| OLS Fit, Compare, Predict | OLSModel / ModelCompareResult — no factory; materialized by fit |
Type help VecNf, help Mat2f, help vec::dot, help mat::*, or help OLSModel in the TUI for the same stack effects inline.
A worked data story that uses these types is the Auto-stats sample (used-car asking price, PRE/F model comparison).
Pattern
3 VecNf.new^ [ 1. 2. 3. ]
2 2 Mat2f.new^ [ 1. 0. 0. 1. ]
mat::matvec // ( A:Mat2f v:VecNf -- w:VecNf )