% softmaxCost.m

function [cost, grad] = mrCost(theta, numOuts, inputSize, lambda, data, labels)

% numOuts - the number of number of outputs 

% inputSize - the size N of the input vector

% lambda - weight decay parameter

% data - the N x M input matrix, where each column data(:, i) corresponds to

% a single test set

% labels - an M x 1 matrix containing the labels corresponding for the input data

%

% Unroll the parameters from theta

theta = reshape(theta, numOuts, inputSize);

numCases = size(data, 2);

%groundTruth = full(sparse(labels, 1:numCases, 1));

cost = 0;

thetagrad = zeros(numOuts, inputSize);

%% ---------- YOUR CODE HERE --------------------------------------

% Instructions: Compute the cost and gradient for softmax regression.

% You need to compute thetagrad and cost.

% The groundTruth matrix might come in handy.

[nfeatures, nsamples] = size(data);

zi = theta * data;

%zi=zi+.0005*randn(size(zi));

hzi = 1./(1+exp(-zi));

temp1 = labels .* log(hzi);

temp2 = (1-labels) .* log(1-hzi);

cost = - sum(sum(temp1+temp2)) ./ nsamples;

cost = cost + sum(sum(theta .^ 2)) .* lambda ./ 2;

temp3 = labels - hzi;

temp4 = temp3 * data';

thetagrad = - temp4 ./ nsamples;

thetagrad = thetagrad + lambda .* theta;

% ------------------------------------------------------------------

% Unroll the gradient matrices into a vector for minFunc

grad = [thetagrad(:)];

end