working through Train() function

nn/main.go

@@ -2,10 +2,12 @@ package nn
 
 import (
 	"fmt"
+	"math/rand"
 	"strings"
 
 	"github.com/go-gota/gota/dataframe"
 	"github.com/gofiber/fiber/v2"
+	"gonum.org/v1/gonum/mat"
 )
 
 type NN struct {
@@ -38,9 +40,45 @@ func NewNN(c *fiber.Ctx) (*NN, error) {
 
 func (nn *NN) Train() {
 	// train test split the data
+	XTrain, XTest, YTrain, YTest := nn.trainTestSplit()
+
+	weights, biases := nn.InitWnB()
 
 	// iterate n times where n = nn.Epochs
 	// use backprop algorithm on each iteration
 	// to fit the model to the data
 
 }
+
+func (nn *NN) InitWnB() {
+	// randomly initialize weights and biases to start
+	inputSize := len(nn.Features)
+	hiddenSize := nn.HiddenSize
+	outputSize := 1 // only predicting one thing for now
+
+	// Initialize weights and biases for the input layer to hidden layer
+	weightsInputHidden := mat.NewDense(inputSize, hiddenSize, nil)
+	weightsInputHidden.Apply(func(_, _ int, v float64) float64 {
+		// Randomly initialize weights with values between -1 and 1
+		return rand.Float64()*2 - 1
+	}, weightsInputHidden)
+
+	biasesHidden := mat.NewVecDense(hiddenSize, nil)
+	biasesHidden.Apply(func(_, _ int, v float64) float64 {
+		// Randomly initialize biases
+		return rand.Float64()
+	}, biasesHidden)
+
+	// Initialize weights and biases for the hidden layer to output layer
+	weightsHiddenOutput := mat.NewDense(hiddenSize, outputSize, nil)
+	weightsHiddenOutput.Apply(func(_, _ int, v float64) float64 {
+		// Randomly initialize weights with values between -1 and 1
+		return rand.Float64()*2 - 1
+	}, weightsHiddenOutput)
+
+	biasesOutput := mat.NewVecDense(outputSize, nil)
+	biasesOutput.Apply(func(_, _ int, v float64) float64 {
+		// Randomly initialize biases
+		return rand.Float64()
+	}, biasesOutput)
+}

nn/split.go

@@ -3,11 +3,11 @@ package nn
 import (
 	"math"
 	"math/rand"
-)
 
-// implement train test split function
+	"github.com/go-gota/gota/dataframe"
+)
 
-func (nn *NN) trainTestSplit() {
+func (nn *NN) trainTestSplit() (dataframe.DataFrame, dataframe.DataFrame, dataframe.DataFrame, dataframe.DataFrame) {
 	// now we split the data into training
 	// and testing based on user specified
 	// nn.TestSize.
@@ -31,8 +31,10 @@ func (nn *NN) trainTestSplit() {
 	// Create the train DataFrame using the trainIndices
 	train := nn.Df.Subset(trainIndices)
 
-	nn.XTrain = train.Select(nn.Features)
-	nn.YTrain = train.Select(nn.Target)
-	nn.XTest = test.Select(nn.Features)
-	nn.YTest = test.Select(nn.Target)
+	XTrain = train.Select(nn.Features)
+	YTrain = train.Select(nn.Target)
+	XTest = test.Select(nn.Features)
+	YTest = test.Select(nn.Target)
+
+	return XTrain, XTest, YTrain, YTest
 }