clear; clc;

% Get Data
gbr1 = imread('images/Banana/1.jpg');
gbr2 = imread('images/Banana/2.jpg');
gbr3 = imread('images/Banana/3.jpg');
gbr4 = imread('images/Banana/4.jpg');
gbr5 = imread('images/Banana/5.jpg');
gbr6 = imread('images/Banana/6.jpg');
gbr7 = imread('images/Banana/7.jpg');
gbr8 = imread('images/Banana/8.jpg');
gbr9 = imread('images/Banana/9.jpg');
gbr10 = imread('images/Banana/10.jpg');
gbr11 = imread('images/Kiwi/1.jpg');
gbr12 = imread('images/Kiwi/2.jpg');
gbr13 = imread('images/Kiwi/3.jpg');
gbr14 = imread('images/Kiwi/4.jpg');
gbr15 = imread('images/Kiwi/5.jpg');
gbr16 = imread('images/Kiwi/6.jpg');
gbr17 = imread('images/Kiwi/7.jpg');
gbr18 = imread('images/Kiwi/8.jpg');
gbr19 = imread('images/Kiwi/9.jpg');
gbr20 = imread('images/Kiwi/10.jpg');
gbr21 = imread('images/Orange/1.jpg');
gbr22 = imread('images/Orange/2.jpg');
gbr23 = imread('images/Orange/3.jpg');
gbr24 = imread('images/Orange/4.jpg');
gbr25 = imread('images/Orange/5.jpg');
gbr26 = imread('images/Orange/6.jpg');
gbr27 = imread('images/Orange/7.jpg');
gbr28 = imread('images/Orange/8.jpg');
gbr29 = imread('images/Orange/9.jpg');
gbr30 = imread('images/Orange/10.jpg');
gbr1=double(rgb2gray(gbr1));
gbr2=double(rgb2gray(gbr2));
gbr3=double(rgb2gray(gbr3));
gbr4=double(rgb2gray(gbr4));
gbr5=double(rgb2gray(gbr5));
gbr6=double(rgb2gray(gbr6));
gbr7=double(rgb2gray(gbr7));
gbr8=double(rgb2gray(gbr8));
gbr9=double(rgb2gray(gbr9));
gbr10=double(rgb2gray(gbr10));
gbr11=double(rgb2gray(gbr11));
gbr12=double(rgb2gray(gbr12));
gbr13=double(rgb2gray(gbr13));
gbr14=double(rgb2gray(gbr14));
gbr15=double(rgb2gray(gbr15));
gbr16=double(rgb2gray(gbr16));
gbr17=double(rgb2gray(gbr17));
gbr18=double(rgb2gray(gbr18));
gbr19=double(rgb2gray(gbr19));
gbr20=double(rgb2gray(gbr20));
gbr21=double(rgb2gray(gbr21));
gbr22=double(rgb2gray(gbr22));
gbr23=double(rgb2gray(gbr23));
gbr24=double(rgb2gray(gbr24));
gbr25=double(rgb2gray(gbr25));
gbr26=double(rgb2gray(gbr26));
gbr27=double(rgb2gray(gbr27));
gbr28=double(rgb2gray(gbr28));
gbr29=double(rgb2gray(gbr29));
gbr30=double(rgb2gray(gbr30));

gbr1=reshape(gbr1',1,[]);
gbr2=reshape(gbr2',1,[]);
gbr3=reshape(gbr3',1,[]);
gbr4=reshape(gbr4',1,[]);
gbr5=reshape(gbr5',1,[]);
gbr6=reshape(gbr6',1,[]);
gbr7=reshape(gbr7',1,[]);
gbr8=reshape(gbr8',1,[]);
gbr9=reshape(gbr9',1,[]);
gbr10=reshape(gbr10',1,[]);
gbr11=reshape(gbr11',1,[]);
gbr12=reshape(gbr12',1,[]);
gbr13=reshape(gbr13',1,[]);
gbr14=reshape(gbr14',1,[]);
gbr15=reshape(gbr15',1,[]);
gbr16=reshape(gbr16',1,[]);
gbr17=reshape(gbr17',1,[]);
gbr18=reshape(gbr18',1,[]);
gbr19=reshape(gbr19',1,[]);
gbr20=reshape(gbr20',1,[]);
gbr21=reshape(gbr21',1,[]);
gbr22=reshape(gbr22',1,[]);
gbr23=reshape(gbr23',1,[]);
gbr24=reshape(gbr24',1,[]);
gbr25=reshape(gbr25',1,[]);
gbr26=reshape(gbr26',1,[]);
gbr27=reshape(gbr27',1,[]);
gbr28=reshape(gbr28',1,[]);
gbr29=reshape(gbr29',1,[]);
gbr30=reshape(gbr30',1,[]);

X = [gbr1, gbr2, gbr3, gbr4, gbr5, gbr6, gbr7, gbr8, gbr9, gbr10, gbr11, gbr12, gbr13, gbr14, gbr15, gbr16, gbr17, gbr18, gbr19, gbr20, gbr21, gbr22, gbr23, gbr24, gbr25, gbr26, gbr27, gbr28, gbr29, gbr30];
[n,p] = size(X);
Xrata = mean(X)

%1. mengurangi setiap variable dengan rata2nya (normalisasi data)
for i = 1:n
    Z(i, :) = X(i, :)-Xrata;
end

%Menghitung covariance matrix
S = cov(X); % varian

%Standardization Data
for i = 1:n
    for j = 1:p
        Zr(i,j) = Z(i,j) / sqrt(S(j,j));
    end
end
%ZR ADALAH HASIL DARI NORMALISASI DATA X (DATA AWAL)

%2. Menghitung cov dari data yang sudah direduksi
Zcov = cov(Zr); % matrix covarian dari variable hasil normalisasi Zr

%3. Menghitung eigen value dan eigen vector dari Zcov
[VA, DA] = eig(Zcov); % VA = Eigen Vector ; DA = Eigen Value

%4. menghitung komponen utama
% Hasil library matlab eig() --> hasil = diagonal utamanya urut dari kecil ke besar
% yang diperlukan dari besar ke kecil
%mengurutkan hasil eigen (DESC)
m = 1
for i = p:-1:1
    DA_urut(m,m) = DA(i,i);
    VA_urut(m,:) = VA(i,:);
    m = m+1;
end

%menghitung komponen utama
for i = 1:p
    eValue(i,1) = DA_urut(i,i);
end

Total = sum(eValue)% total varians dalam data asli yang dapat dijelaskan oleh seluruh komponen utama.

% Proporsi kumulatif = mengambil data seberapa persen secara kumulatif yang
% dianggap perlu
PK = 0;
komponen = 0;%jumlah komponen utama yang dipertahankan

for j = 1:p
    proporsiValue(j,:) = eValue(j,1) / Total;%mengukur seberapa besar kontribusi eigenvalue ke-j terhadap varians total data.
    if(PK < 0.9) % Nilai PK tergantung peneliti --> 0.8
        PK = PK + proporsiValue(j, :);
        komponen = komponen + 1;
    end
end

% Mengambil eigen vector dari komponen yang terpilih
VA_Komponen = VA_urut(:,1:komponen)
% W = variable komponen utama / principal component variable
W = Z * VA_Komponen % W ADALAH FINAL DATA YANG DISEBUTKAN DI PDF

plot3(W(:,1), W(:,2), W(:,3), '*')        % PK < 0.9
 %plot(W(:,1), W(:,2), '*')                   % PK < 0.8
grid on