% This script plots three QQ-normal plots, for a normal, gamma, % and t distribution. % Set up the figure. figure set(gcf, 'Position', [0, 200, 1900 500]) fontSize = 18; markerSize = 3; % Set simulation parameters. N = 200 % ProbDistUnivParam is a Matlab function to create a probability % distribution object. It's used in the qq plotting function. standardNormal = ProbDistUnivParam('normal', [0, 1]); % Draw the samples for the empircal CDFs. normalSample = normrnd(0, 1, N, 1); chiSquaredSample = chi2rnd(4, N, 1); tSample = trnd(3, N, 1); % Plot the QQ normal for a normal sample. subplot(1, 3, 1) qqNormal = qqplot(normalSample, standardNormal); set(qqNormal, 'Color', 'w', 'MarkerSize', markerSize, 'MarkerEdgeColor', 'k') set(gca, 'XTick', -4:1:4, 'YTick', -4:1:4, 'TickDir', 'out') xlabel('Theoretical Quantiles', 'FontSize', fontSize) ylabel('') title('') % Plot the QQ normal for a chi squared sample. subplot(1, 3, 2) qqChiSquared = qqplot(chiSquaredSample, standardNormal); set(qqChiSquared, 'Color', 'w', 'MarkerSize', markerSize, 'MarkerEdgeColor', 'k') set(gca, 'XTick', -4:1:4, 'YTick', 0:2:18, 'TickDir', 'out') ylim([0, ceil(max(chiSquaredSample))]) xlabel('Theoretical Quantiles', 'FontSize', fontSize) ylabel('') title('') % Plot the QQ normal for a t sample. subplot(1, 3, 3) qqT = qqplot(tSample, standardNormal); set(qqT, 'Color', 'w', 'MarkerSize', markerSize, 'MarkerEdgeColor', 'k') set(gca, 'XTick', -4:1:4, 'YTick', -8:2:8, 'TickDir', 'out') xlabel('Theoretical Quantiles', 'FontSize', fontSize) ylabel('') title('')