mapping activation arg to the actual function

nn/backprop.go

@@ -0,0 +1,8 @@
+package nn
+
+func (nn *NN) Backprop() {
+
+	for i := 0; i < nn.Epochs; i++ {
+	}
+
+}

nn/main.go

@@ -10,24 +10,27 @@ import (
 )
 
 type NN struct {
-	CSVData        string   `json:"csv_data"`
-	Features       []string `json:"features"`
-	Target         string   `json:"target"`
-	Epochs         int      `json:"epochs"`
-	HiddenSize     int      `json:"hidden_size"`
-	LearningRate   float64  `json:"learning_rate"`
-	ActivationFunc string   `json:"activation"`
-	TestSize       float64  `json:"test_size"`
+	// attributes set by request
+	CSVData      string   `json:"csv_data"`
+	Features     []string `json:"features"`
+	Target       string   `json:"target"`
+	Epochs       int      `json:"epochs"`
+	HiddenSize   int      `json:"hidden_size"`
+	LearningRate float64  `json:"learning_rate"`
+	Activation   string   `json:"activation"`
+	TestSize     float64  `json:"test_size"`
 
-	Df     *dataframe.DataFrame
-	XTrain dataframe.DataFrame
-	YTrain dataframe.DataFrame
-	XTest  dataframe.DataFrame
-	YTest  dataframe.DataFrame
-	Wh     [][]float64
-	Bh     []float64
-	Wo     [][]float64
-	Bo     []float64
+	// attributes set after args above are parsed
+	ActivationFunc *func(float64) float64
+	Df             *dataframe.DataFrame
+	XTrain         *dataframe.DataFrame
+	YTrain         *dataframe.DataFrame
+	XTest          *dataframe.DataFrame
+	YTest          *dataframe.DataFrame
+	Wh             *[][]float64
+	Bh             *[]float64
+	Wo             *[][]float64
+	Bo             *[]float64
 }
 
 func NewNN(c *fiber.Ctx) (*NN, error) {
@@ -37,24 +40,13 @@ func NewNN(c *fiber.Ctx) (*NN, error) {
 		return nil, fmt.Errorf("invalid JSON data: %v", err)
 	}
 	df := dataframe.ReadCSV(strings.NewReader(newNN.CSVData))
+	activation := ActivationMap[newNN.Activation]
+
 	newNN.Df = &df
+	newNN.ActivationFunc = &activation
 	return newNN, nil
 }
 
-func (nn *NN) Train() {
-	// train test split the data
-	_, _, _, _ = nn.trainTestSplit()
-
-	nn.InitWnB()
-
-	// iterate n times where n = nn.Epochs
-	// use backprop algorithm on each iteration
-	// to fit the model to the data
-	for i := 0; i < nn.Epochs; i++ {
-	}
-
-}
-
 func (nn *NN) InitWnB() {
 	// randomly initialize weights and biases to start
 	inputSize := len(nn.Features)
@@ -89,8 +81,8 @@ func (nn *NN) InitWnB() {
 		bo[i] = rand.Float64() - 0.5
 	}
 
-	nn.Wh = wh
-	nn.Bh = bh
-	nn.Wo = wo
-	nn.Bo = bo
+	nn.Wh = &wh
+	nn.Bh = &bh
+	nn.Wo = &wo
+	nn.Bo = &bo
 }

nn/split.go

@@ -3,11 +3,9 @@ package nn
 import (
 	"math"
 	"math/rand"
-
-	"github.com/go-gota/gota/dataframe"
 )
 
-func (nn *NN) trainTestSplit() (dataframe.DataFrame, dataframe.DataFrame, dataframe.DataFrame, dataframe.DataFrame) {
+func (nn *NN) TrainTestSplit() {
 	// now we split the data into training
 	// and testing based on user specified
 	// nn.TestSize.
@@ -36,5 +34,9 @@ func (nn *NN) trainTestSplit() (dataframe.DataFrame, dataframe.DataFrame, datafr
 	XTest := test.Select(nn.Features)
 	YTest := test.Select(nn.Target)
 
-	return XTrain, XTest, YTrain, YTest
+	nn.XTrain = &XTrain
+	nn.YTrain = &YTrain
+	nn.XTest = &XTest
+	nn.YTest = &YTest
+
 }

nn/subset.go

@@ -3,4 +3,4 @@ package nn
 // subset the data frame into just the
 // features and target that the user specify
 
-func subset() {}
+func Subset() {}

nn/train.go

@@ -0,0 +1,13 @@
+package nn
+
+func (nn *NN) Train() {
+	nn.InitWnB()
+	nn.TrainTestSplit()
+
+	// iterate n times where n = nn.Epochs
+	// use backprop algorithm on each iteration
+	// to fit the model to the data
+	for i := 0; i < nn.Epochs; i++ {
+	}
+
+}

server.go

@@ -1,8 +1,6 @@
 package main
 
 import (
-	"fmt"
-
 	"github.com/Jensen-holm/ml-from-scratch/nn"
 	"github.com/gofiber/fiber/v2"
 )
@@ -13,7 +11,6 @@ func main() {
 	// eventually we might want to add a key to this endpoint
 	// that we will be able to validate.
 	app.Post("/neural-network", func(c *fiber.Ctx) error {
-
 		nn, err := nn.NewNN(c)
 		if err != nil {
 			return c.Status(fiber.StatusBadRequest).JSON(fiber.Map{
@@ -22,7 +19,6 @@ func main() {
 		}
 
 		nn.Train()
-		fmt.Println(nn.Wo)
 
 		return c.SendString("No error")
 	})