linear-regression ( X y lr n-iters -- )

Vocabulary

tensors.demos

Inputs

X	an object
y	an object
lr	an object
n-iters	an object

Outputs

None

Definition

USING: accessors arrays io kernel math.parser prettyprint
sequences tensors tensors.demos.private ;

IN: tensors.demos

:: linear-regression ( X y lr n-iters -- )
    X normalize y shape>> first 1 2array ones swap 2array
    hstack :> X-norm X-norm shape>> second 1 2array
    zeros :> params X-norm y params compute-cost
    number>string "The initial cost is " swap append print
    X-norm y params lr n-iters gradient-descent
    "The optimal parameters are " print .
    last number>string "The final cost was " swap append print ;