No subject

the space of nx1 matrices are isomorphic. This is (ab)used
so often in formulas, that I find it convenient to represent
column vectors as matrices. For example: (matrix* (T x) x)
will work for a column matrix x.

#lang racket
(require math)
(define T matrix-transpose)

(define (fit x y)
  (matrix-solve (matrix* (T x) x) (matrix* (T x) y)))