+ linear regression solver

Makefile

@@ -1,5 +1,6 @@
 dev:
 	GOOS=js GOARCH=wasm tinygo build -o stat/mod.wasm -no-debug ./stat/main.go
+	GOOS=js GOARCH=wasm tinygo build -o regr/mod.wasm  ./regr/main.go
 	GOOS=js GOARCH=wasm go build -o plot/mod.wasm ./plot/main.go
 
 prod:

regr/main.go

@@ -11,7 +11,7 @@ import (
 
 func InitRegrExports(this js.Value, args []js.Value) interface{} {
 	exports := args[0]
-	exports.Set("ABCD", js.FuncOf(src.ABCD))
+	exports.Set("Linear", js.FuncOf(src.NewLinearRegressionJS))
 	return nil
 }
 

regr/mod.ts

@@ -1,5 +1,4 @@
 import "../lib/wasm_tinygo.js";
-import type { Stat } from "./types.ts";
 
 // @ts-expect-error: no types
 const go = new Go();
@@ -13,9 +12,9 @@ const wasm = wasmMmodule.instance;
 
 go.run(wasm);
 
-const _exports = {} as Record<string, (...args: unknown[]) => unknown> & Stat;
+const _exports = {} as Record<string, (...args: unknown[]) => unknown>;
 
 // @ts-ignore: no types
-__InitStatExports(_exports);
+__InitRegrExports(_exports);
 
 export default _exports;

regr/src/Linear.go

@@ -1,16 +0,0 @@
-//go:build js && wasm
-// +build js,wasm
-
-package src
-
-import "syscall/js"
-
-// ref: mat.Dense
-// fit: solve least squares
-// predict: predict y from x
-// save: save model
-// load: load model
-// note: separate wasm/js glue
-func ABCD(this js.Value, args []js.Value) interface{} {
-	return nil
-}

regr/src/LinearRegression.go

@@ -0,0 +1,59 @@
+//go:build js && wasm
+// +build js,wasm
+
+package src
+
+import (
+	"syscall/js"
+
+	"gonum.org/v1/gonum/mat"
+)
+
+// ref: mat.Dense
+// fit: solve least squares
+// predict: predict y from x
+// save: save model
+// load: load model
+// note: separate wasm/js glue
+
+type LinearRegression struct {
+	Coef *mat.Dense
+}
+
+func (reg *LinearRegression) Fit(X, Y [][]float64) error {
+	XDense := Array2DToDense(X)
+	YDense := Array2DToDense(Y)
+	reg.Coef = new(mat.Dense)
+	reg.Coef.Solve(XDense, YDense)
+	return nil
+}
+
+func (reg *LinearRegression) Predict(X [][]float64) ([]float64, error) {
+	YDense := new(mat.Dense)
+	YDense.Mul(Array2DToDense(X), reg.Coef)
+	return YDense.RawMatrix().Data, nil
+}
+
+func (l *LinearRegression) Save() ([]byte, error) {
+	return nil, nil
+}
+
+func (l *LinearRegression) Load(data []byte) error {
+	return nil
+}
+
+func NewLinearRegressionJS(this js.Value, args []js.Value) interface{} {
+	reg := new(LinearRegression)
+	obj := js.Global().Get("Object").New()
+	obj.Set("fit", js.FuncOf(func(this js.Value, args []js.Value) interface{} {
+		X := JSFloatArray2D(args[0])
+		Y := JSFloatArray2D(args[1])
+		return reg.Fit(X, Y)
+	}))
+	obj.Set("predict", js.FuncOf(func(this js.Value, args []js.Value) interface{} {
+		X := JSFloatArray2D(args[0])
+		Y, _ := reg.Predict(X)
+		return ToJSArray(Y)
+	}))
+	return obj
+}

regr/src/utils.go

@@ -0,0 +1,39 @@
+//go:build js && wasm
+// +build js,wasm
+
+package src
+
+import (
+	"syscall/js"
+
+	"gonum.org/v1/gonum/mat"
+)
+
+func Array2DToDense(X [][]float64) *mat.Dense {
+	dense := mat.NewDense(len(X), len(X[0]), nil)
+	for i, row := range X {
+		dense.SetRow(i, row)
+	}
+	return dense
+}
+
+func JSFloatArray2D(arg js.Value) [][]float64 {
+	arr := make([][]float64, arg.Length())
+	for i := 0; i < len(arr); i++ {
+		arr[i] = make([]float64, arg.Index(i).Length())
+	}
+	for i := 0; i < len(arr); i++ {
+		for j := 0; j < arg.Index(i).Length(); j++ {
+			arr[i][j] = arg.Index(i).Index(j).Float()
+		}
+	}
+	return arr
+}
+
+func ToJSArray[T any](arr []T) []interface{} {
+	jsArr := make([]interface{}, len(arr))
+	for i, v := range arr {
+		jsArr[i] = v
+	}
+	return jsArr
+}