NeuralNetwork/neuralnetwork/main.go

package main

import (
	"fmt"
	"log"
	"os"
	"time"

	neuralnetwork "./neuralnetworkbase"
	remotecontrol "./remotecontrol"
	teach "./teach"
)

func main() {
	sizes := []int{13, 8, 12, 3}
	nn, _ := neuralnetwork.NewNeuralNetwork(sizes, neuralnetwork.NewRPropInitializer(neuralnetwork.RPropConfig{
		NuPlus:   1.2,
		NuMinus:  0.5,
		DeltaMax: 50.0,
		DeltaMin: 0.000001,
	}))

	rc := &remotecontrol.RemoteControl{}
	nn.SetStateWatcher(rc)
	// inFile, err := os.Open("./networkstate")
	// if err != nil {
	// 	log.Fatal(err)
	// }
	// defer inFile.Close()
	// nn.LoadState(inFile)

	// nn, _ := neuralnetwork.NewNeuralNetwork(sizes, neuralnetwork.NewBackPropInitializer(0.1))

	// for i := 0; i < nn.Count; i++ {
	// 	if i > 0 {
	// 		fmt.Printf("Weights before:\n%v\n\n", mat.Formatted(nn.Weights[i], mat.Prefix(""), mat.Excerpt(0)))
	// 		fmt.Printf("Biases before:\n%v\n\n", mat.Formatted(nn.Biases[i], mat.Prefix(""), mat.Excerpt(0)))
	// 		fmt.Printf("Z before:\n%v\n\n", mat.Formatted(nn.Z[i], mat.Prefix(""), mat.Excerpt(0)))
	// 	}
	// 	fmt.Printf("A before:\n%v\n\n", mat.Formatted(nn.A[i], mat.Prefix(""), mat.Excerpt(0)))
	// }

	go func() {
		// teacher := teach.NewMNISTReader("./minst.data", "./mnist.labels")
		teacher := teach.NewTextDataReader("wine.data", 5)
		nn.Teach(teacher, 500)

		// for i := 0; i < nn.Count; i++ {
		// 	if i > 0 {
		// 		fmt.Printf("Weights after:\n%v\n\n", mat.Formatted(nn.Weights[i], mat.Prefix(""), mat.Excerpt(0)))
		// 		fmt.Printf("Biases after:\n%v\n\n", mat.Formatted(nn.Biases[i], mat.Prefix(""), mat.Excerpt(0)))
		// 		fmt.Printf("Z after:\n%v\n\n", mat.Formatted(nn.Z[i], mat.Prefix(""), mat.Excerpt(0)))
		// 	}
		// 	fmt.Printf("A after:\n%v\n\n", mat.Formatted(nn.A[i], mat.Prefix(""), mat.Excerpt(0)))
		// }

		outFile, err := os.OpenFile("./data", os.O_CREATE|os.O_TRUNC|os.O_WRONLY, 0666)
		if err != nil {
			log.Fatal(err)
		}
		defer outFile.Close()
		nn.SaveState(outFile)
		outFile.Close()

		time.Sleep(5 * time.Second)
		failCount := 0
		teacher.Reset()
		for true {
			if !teacher.NextValidator() {
				fmt.Printf("Fail count: %v\n\n", failCount)
				failCount = 0
				teacher.Reset()
			}
			dataSet, expect := teacher.GetValidator()
			index, _ := nn.Predict(dataSet)
			//TODO: remove this is not used for visualization
			time.Sleep(400 * time.Millisecond)
			if expect.At(index, 0) != 1.0 {
				failCount++
				// fmt.Printf("Fail: %v, %v\n\n", teacher.ValidationIndex(), expect.At(index, 0))
			}
		}
	}()

	// nn = &neuralnetwork.NeuralNetwork{}
	// inFile, err := os.Open("./data")
	// if err != nil {
	// 	log.Fatal(err)
	// }
	// defer inFile.Close()
	// nn.LoadState(inFile)
	// inFile.Close()

	// failCount = 0
	// teacher.Reset()
	// for teacher.NextValidator() {
	// 	dataSet, expect := teacher.GetValidator()
	// 	index, _ := nn.Predict(dataSet)
	// 	if expect.At(index, 0) != 1.0 {
	// 		failCount++
	// 		// fmt.Printf("Fail: %v, %v\n\n", teacher.ValidationIndex(), expect.At(index, 0))
	// 	}
	// }

	// fmt.Printf("Fail count: %v\n\n", failCount)
	rc.Run()
}