%% This loads our data
[X, y] = load_data_ex2();
y

%% Normalise and initialize.
[X, mean_vec, std_vec] = normalise_features(X);

%after normalising we add the bias
X = [ones(size(X, 1), 1), X];

%initialise theta
theta = [0.0, 0.0, 0.0];
% Learning rate
alpha = 0.2;
% Number of iterations for calculating theta values in gradient descent
iterations = 100;

% Perform gradient descent on data to calculate the values for theta that make
% the best hypothesis function.
t = gradient_descent(X, y, theta, alpha, iterations, false)

% Declare new pair of input features with which to calculate a predicted output
% value.
input_size = 3000.0;
input_bedrooms = 4.0;

% Normalize features
is_norm = (input_size - mean_vec(1)) / std_vec(1)
ib_norm = (input_bedrooms - mean_vec(2)) / std_vec(2)

% Use values for theta generated by gradient descent to predict the output
prediction = t(1) + t(2)*is_norm + t(3)*ib_norm;
% Round and print output
round(prediction)

disp 'Press enter to exit!';
pause;