// Ann.cpp : Defines the entry point for the console application.
//

#include "stdafx.h"
# include <cmath>
#include <math.h>
#include <vector>
#include <iostream>
#include <iomanip>
#include <fstream>
#include <string>
#include <array>
using namespace std;
//Generating random number: either 0 or 1, uniform distribution, for XOR operation. Can remove later if using data from files.
int randint();
double f(double x);
double f_deriv(double x);
double gL(double a, double z, double t);
double w_gradient(int layer_id, int w_i, int w_j, double *a_arr, int *s, double *gll);
double delta_w(double grad, double dw);
void calc_gjl(double *a_arr, double *z_arr, double *t_arr, double *w_arr, int *s, int *sw, int L, int *l, double *gll);
const double ETA = 0.1;
const double ALPHA = 0.15;
void file(string filename, double *a, int dydis);

struct Topology
{
    std::vector<int> l;//kiekiai sluoksnyje
} topolygy;

struct Sample
{
    double input[2];
    double output[2];
    //double * input=new double[topolygy.l.at(0)];
    //double * output=new double[topolygy.l.at(topolygy.l.size()-1)];
    //Sample(double a[],double b[]) {
    //  input = a;
    //  output = b;
    //};
    //temp*******
    string ToString() {
        string str;
        str = "input: " + to_string(input[0]) + " " + to_string(input[1]) + " output: " + to_string(output[0]) + " " + to_string(output[1]) + "\n";
        return str;
    }
};
class Data
{
public:
    int getNumberOfInputs() { return inputs; }
    int getNumberOfOutputs() { return outputs; }

    double * getInput(int index);

    double * getOutput(int index);

    int getNumberOfSamples() { return samples; }

    void addSample(Sample sample);

    void setSizes(int input_size, int output_size);

protected:
    std::vector<Sample> data;
    int inputs;
    int outputs;
    int samples = 0;
};

class XOR : public Data
{
public:
    void generate(int n);

    XOR()
    {
        inputs = 2;
        outputs = 2;
        samples = 0;
    }
    void printInputs(int index);

    void printOutputs(int index);
};

class AnnBase {
public:
    virtual void prepare(Topology top) = 0;
    virtual void init(Topology top, double w_arr_1[]) = 0;
    virtual void train(double *a, double *b) = 0;
    virtual void feedForward(double *a, double *b) = 0;
    virtual void destroy() = 0;
private:
    virtual void calc_feedForward() = 0;
};

class AnnSerialDBL : public AnnBase {
public:
    void prepare(Topology top);
    void init(Topology top, double w_arr_1[]);
    void train(double *a, double *b);
    void feedForward(double *a, double *b);
    void destroy();
private:
    void calc_feedForward();
public:
    int z_count;//temp var to keep the length of z, so z could be reset for calcs.
    int inputs;
    int outputs;
    int L;
    int * l;
    int * s;
    double * a_arr;
    double * z_arr;
    int * W;
    int * sw;
    double * w_arr;
    double * dw_arr;
    double * t_arr;
    double * gjl;
};
void print_all(AnnSerialDBL SerialDBL, int sum, int mult, int i);
void read_W(string filename, double *w_arr);
void calc_sizes(int &sum, int &mult, Topology top);
int main()
{
    //15+24+10  49
    topolygy.l.push_back(2);
    topolygy.l.push_back(5);
    topolygy.l.push_back(4);
    //topolygy.l.push_back(7);
    topolygy.l.push_back(2);
    AnnSerialDBL SerialDBL;


    int sum = 0;
    int mult = 0;
    calc_sizes(sum, mult, topolygy);

    SerialDBL.prepare(topolygy);

    double * a1=new double[mult];
    read_W("W_ARR.txt", a1);

    SerialDBL.init(topolygy, a1);
    delete[] a1;
    a1 = NULL;

    print_all(SerialDBL, sum, mult, 0);

    XOR xo;
    xo.generate(100000);
    SerialDBL.train(xo.getInput(0), xo.getOutput(0));
    print_all(SerialDBL, sum, mult, 1);
    for (int i = 1; i < xo.getNumberOfSamples(); i++) {
        SerialDBL.train(xo.getInput(i), xo.getOutput(i));
    }

    //Checking results(all combinations 0 and 1)
    for (double i = 0; i < 2; i++) {
        for (double j = 0; j < 2; j++) {
            double input[] = { i ,j };
            double output[] = { 0,0 };
            SerialDBL.feedForward(input, output);

            cout << endl << "input : " << input[0] << "   " << input[1] << endl;
            cout << endl << "output: " << output[0] << "   " << output[1] << endl << endl;
            cout << "---------------------------------------------------" << endl;
        }
    }

    //Checking results(all combinations 0 and 1)
    for (double i = 0; i < 0; i++) {
        double input[] = { randint()*1.0, randint()*1.0 };
        double output[] = { 0,0 };
        SerialDBL.feedForward(input, output);

        cout << endl << "input : " << input[0] << "   " << input[1] << endl;
        cout << endl << "output: " << output[0] << "   " << output[1] << endl << endl;
        cout << "---------------------------------------------------" << endl;
    }

    print_all(SerialDBL, sum, mult, 2);


    SerialDBL.destroy();

    int a;
    cin >> a;

    return 0;
}
//returns random int, either 0 or 1
int randint() {
    double r = ((double)rand() / (RAND_MAX));
    int a = 0;
    if (r > 0.5) {
        a = 1;
    }
    else
    {
        a = 0;
    }
    return a;
}

double f(double x) {
    double y = 1 + exp(-x);
    //temp*********************
    if (y == 0) {
        cout << "Error 1";
    }
    if ((y - 1) == 0) {
        //cout << "Error 2";
    }
    //temp**********************
    return 1 / y;
}

double f_deriv(double x) {
    //Temp**********
    double y = pow((1 + exp(-x)), 2);
    double z = exp(-x);
    if (y == 0) {
        cout << "Error 3";
    }
    if (z == 0) {
    //  cout << "Error 4";
    }
    //temp**********************
    return exp(-x) / pow((1 + exp(-x)), 2);
}

double gL(double a, double z, double t) {
    double w = f_deriv(z) * (a - t);
    //cout << "z: " << z << " a: " << a << " t: " << t << endl;
    return w;
}

double w_gradient(int layer_id, int w_i, int w_j, double *a_arr, int *s, double *gll) {
    return a_arr[s[layer_id] + w_i] * gll[s[layer_id+1] + w_j];
}

double delta_w(double grad, double dw) {
    return (-ETA)*grad + ALPHA*dw;
}

void calc_gjl(double *a_arr, double *z_arr, double *t_arr, double *w_arr, int *s, int *sw, int L, int *l, double *gll)
{
    for (int i = L-2; i >= 0; i--) {
        for (int j = 0; j < l[i + 1] - 1; j++) {
            if (L - 2 == i) {
                gll[s[i + 1] + j] = gL(a_arr[s[i + 1] + j], z_arr[s[i + 1] + j], t_arr[j]);
            }
            else {
                gll[s[i + 1] + j] = f_deriv(z_arr[s[i + 1] + j]);
                double sum = 0;
                for (int k = 0; k < l[i+2]-1; k++) {
                    sum += w_arr[sw[i + 1] + k*(l[i + 2] - 1) + j] * gll[s[i + 2] + j];
                }
                gll[s[i + 1] + j] *= sum;
            }
        }
    }
}

//*********
double * Data::getInput(int index)
{
    return data[index].input;
}

double * Data::getOutput(int index)
{
    return data[index].output;
}

void Data::addSample(Sample sample)
{
    data.push_back(sample);
    samples++;
    //cout << sample.ToString();
}

void Data::setSizes(int input_size, int output_size)
{
    inputs = input_size;
    outputs = output_size;
}

//****************
void XOR::generate(int n)
{
    for (int i = 0; i < n / 4; i++)
    {
        //double input1 = randint();
        //double input2 = randint();
        for (double j = 0; j < 2; j++) {
            for (double k = 0; k < 2; k++) {
                double output1 = j == k;
                double output2 = j != k;
                double input[] = { j,k };
                double output[] = { output1, output2 };
                //Sample s = { input,output };
                //addSample({ input,output });
                addSample({ {j,k} ,{output1,output2} });
            }
        }

    }
}

void XOR::printInputs(int index)
{
    cout << index << " index inputs: " << endl;
    for (int i = 0; i < inputs; i++)
    {
        cout << getInput(index)[i] << " ";
    }
    cout << endl;
}

void XOR::printOutputs(int index)
{
    cout << index << " index outputs: " << endl;
    for (int i = 0; i < outputs; i++)
    {
        cout << fixed << setprecision(2) << data[index].output[i] << " ";
    }
    cout << endl;
}

//****************
void AnnSerialDBL::prepare(Topology top)
{
    inputs = top.l.at(0);
    outputs = top.l.at(top.l.size() - 1);

    l = new int[top.l.size()];
    s = new int[top.l.size()];

    int sum = 0;
    int mult = 0;
    for (int i = 0; i < top.l.size(); i++) {
        sum += top.l.at(i) + 1;
    }
    z_count = sum;
    for (int i = 0; i < top.l.size() - 1; i++) {
        mult += (top.l.at(i) + 1)*top.l.at(i + 1);
    }
    a_arr = new double[sum];
    z_arr = new double[sum];

    W = new int[top.l.size()];
    sw = new int[top.l.size()];

    w_arr = new double[mult];
    dw_arr = new double[mult];

    t_arr = new double[top.l.at(top.l.size() - 1)];

    gjl = new double[sum];
}

void AnnSerialDBL::init(Topology top, double w_arr_1[] = NULL)
{
    L = top.l.size();
    //Neuronu kiekiai sluoksnyje
    for (int i = 0; i < top.l.size(); i++) {
        l[i] = top.l.at(i) + 1;
    }

    //Sluoksniu pradzios indeksai
    for (int i = 0; i < top.l.size(); i++) {
        s[i] = 0;
        for (int j = i; j > 0; j--) {
            s[i] += l[j - 1];
        }
    }

    //Bias neuronai
    for (int i = 0; i < top.l.size() - 1; i++) {
        a_arr[s[i + 1] - 1] = 1;
    }

    //Svoriu kiekiai l-ame sluoksnyje
    for (int i = 0; i < top.l.size() - 1; i++) {
        W[i] = l[i] * (l[i + 1] - 1);
        //cout << "Svoriu sk: " << W[i] << " Pradzios index: ";
        sw[i] = 0;
        if (i != 0) {
            for (int j = 0; j < i; j++) {
                sw[i] += W[j];
            }
        }
        if (w_arr_1 == NULL) {
            for (int j = 0; j < W[i]; j++) {
                w_arr[sw[i] + j] = (double)rand() / double(RAND_MAX);
                //cout << w_arr[sw[i] + j]<< endl;
                dw_arr[sw[i] + j] = 0;
            }
        }
        else {
            //w_arr = w_arr_1; //ar reikia pokycius issisaugoti irgi?
            for (int j = 0; j < W[i]; j++) {
                w_arr[sw[i] + j] = w_arr_1[sw[i] + j];
                //cout << w_arr[sw[i] + j]<< endl;
                dw_arr[sw[i] + j] = 0;
            }
        }

        //cout << sw[i] << " " << endl;
    }
}

void AnnSerialDBL::train(double *a, double *b)
{
    for (int i = 0; i < inputs; i++) {
        a_arr[i] = a[i];
    }

    for (int j = 0; j < z_count; j++) {
        z_arr[j] = 0;
    }

    calc_feedForward();

    for (int i = 0; i < outputs; i++) {
        t_arr[i] = b[i];
    }
    calc_gjl(a_arr, z_arr, t_arr, w_arr, s, sw, L,l, gjl);

    //back propogation:
    for (int i = L - 2; i >= 0; i--) {//per sluoksnius
        for (int j = 0; j < l[i]; j++) {//per neuronus
            for (int k = 0; k < l[i + 1] - 1; k++) {//per kito sluoksnio neuronus
                dw_arr[sw[i] + k + j*(l[i + 1] - 1)] = delta_w(w_gradient(i, j, k, a_arr, s,gjl), dw_arr[sw[i] + k + j*(l[i + 1] - 1)]);
                w_arr[sw[i] + k + j*(l[i + 1] - 1)] += dw_arr[sw[i] + k + j*(l[i + 1] - 1)];
            }
        }
    }
}

void AnnSerialDBL::feedForward(double *a, double *b)
{
    for (int i = 0; i < inputs; i++) {
        a_arr[i] = a[i];
    }

    for (int j = 0; j < z_count; j++) {
        z_arr[j] = 0;
    }

    calc_feedForward();

    double max = 0;
    int index = 0;

    for (int i = 0; i<outputs; i++) {
        cout << " a reiksmes: " << a_arr[s[L - 1] + i] << endl;
        if (max < a_arr[s[L - 1] + i]) {
            max = a_arr[s[L - 1] + i];
            index = i;
        }
    }
    for (int i = 0; i < outputs; i++) {
        if (i == index) {
            b[i] = 1;
        }
        else {
            b[i] = 0;
        }
    }
}

void AnnSerialDBL::calc_feedForward()
{
    for (int i = 0; i < L - 1; i++) {//per sluoksnius einu+
        for (int j = 0; j < l[i]; j++) { //kiek neuronu sluoksnyje+
            for (int k = 0; k < l[i + 1] - 1; k++) {//per sekancio sluoksnio z+
                z_arr[s[i + 1] + k] += w_arr[sw[i] + k + j*(l[i + 1] - 1)] * a_arr[s[i] + j];
                //  cout << "w: "<< w_arr[sw[i] + k + j*(l[i + 1] - 1)] << endl;
                //  cout << "a: " << a_arr[s[i] + j] << endl;
                //  cout << "z reiksmes: " << z_arr[s[i+1] + k] << endl;
                //  cout << endl;
            }
        }
        for (int k = 0; k < l[i + 1] - 1; k++) {//per sekancio sluoksnio z
            a_arr[s[i + 1] + k] = f(z_arr[s[i + 1] + k]);
            //  cout << s[i + 1] + k << " a reiksmes: " << a_arr[s[i + 1] + k] << endl;
        }
    }
}

void AnnSerialDBL::destroy()
{
    delete[] l;
    l = NULL;
    delete[] s;
    s = NULL;

    delete[] a_arr;
    a_arr = NULL;
    delete[] z_arr;
    z_arr = NULL;

    delete[] W;
    W = NULL;
    delete[] sw;
    sw = NULL;

    delete[] w_arr;
    w_arr = NULL;
    delete[] dw_arr;
    dw_arr = NULL;

    delete[] t_arr;
    t_arr = NULL;

    delete[] gjl;
    gjl = NULL;
}

void file(string filename, double *a,int dydis) {
    ofstream myfile;
    myfile.open(filename);
    myfile << filename << ";" << endl;
    for (int i = 0; i < dydis; i++) {
        myfile << a[i] << ";" << endl;
    }
    myfile.close();
}

void print_all(AnnSerialDBL SerialDBL, int sum, int mult,int i) {
    file("(" + to_string(5 * i + 1) + ")a_arr.csv", SerialDBL.a_arr, sum);
    file("(" + to_string(5 * i + 2) + ")z_arr.csv", SerialDBL.z_arr, sum);
    file("(" + to_string(5 * i + 3) + ")w_arr.csv", SerialDBL.w_arr, mult);
    file("(" + to_string(5 * i + 4) + ")dw_arr.csv", SerialDBL.dw_arr, mult);
    file("(" + to_string(5 * i + 5) + ")g_arr.csv", SerialDBL.gjl, sum);
}

void read_W(string filename,double *w_arr) {
    ifstream myReadFile;
    myReadFile.open(filename);
    string a;
    if (myReadFile.is_open()) {
        int i = 0;
        while (!myReadFile.eof()) {
            myReadFile >> a;
            w_arr[i++] = stod(a);
        }
    }
    myReadFile.close();
}

void calc_sizes(int &sum, int &mult, Topology top) {
    for (int i = 0; i < topolygy.l.size(); i++) {
        sum += topolygy.l.at(i) + 1;
    }
    for (int i = 0; i < topolygy.l.size() - 1; i++) {
        mult += (topolygy.l.at(i) + 1)*topolygy.l.at(i + 1);
    }
}