Beyond Matlab Plot Functionality, Example 2

The following example of a histogram plot uses filter to fill figure elements with pixel images. The histogram bars will drop a shadow. The axis labels will be rotated by 90 degree to improve readability. All these additional effects are not (yet) supported by Matlab. Thus, the figure shown in Matlab will not contain any effects. Only after printing the figure with plot2svg the effects become visible.

function demo_svg_water
% Create a new figure and remember the figure handle
fig = figure;
% Set default font and font size
set(fig, 'DefaultAxesFontName', 'Arial')
set(fig, 'DefaultAxesFontSize', 16)
% This is the data [year water_consumption]
data = [
1980 230
1982 265
1984 265
1986 266
1988 262
1990 259
1992 277
1994 247
1996 238
1998 250
2000 250
2002 235
2004 233
2006 228
% Let's plot the data
hold on
s = bar(data(:, 1), data(:, 2));
% We do not want to apply the filter to the edge of the bars.
Therefore, we plot them in addition on top
h = bar(data(:, 1), data(:, 2));
set(h, 'FaceColor', 'none');
set(h, 'EdgeColor', 'black');
% Add the text inside each bar with correct rotation and alignment
t = text(data(:, 1), data(:, 2) - 10, num2str(data(:, 2)));
set(t, 'Rotation', -90);
set(t, 'FontSize', 16);
set(t, 'FontName', 'Arial');
set(t, 'VerticalAlignment', 'middle');
set(t, 'FontWeight', 'bold');
% Adding labels and tick marks
axis([1979 2008 0 400])
set(gca, 'XTick', data(:, 1));
set(gca, 'XTickLabel', num2str(data(:, 1)));
title('Water Consumption per Person and Year')
ylabel('Water consumption [l]')
grid on
box on
set(gca, 'Position', [0.13 0.17 0.80 0.72]);
% Now we add the filters
% The bounding box with extension axes makes sure that we cover the full
% axis region with the background images. Due to the shadow we have to

% define an overlap region of 12px. Otherwise, distortions at the border of
% the axis reagion may be visible.
svgBoundingBox(s, 'axes', 12, 'off');
% Now we add the image. To keep the aspect ratio we use 'xMidYMidslice'.
% This setting centers the picture for both x and y. The keyword 'slice'

% makes sure that the picture is scaled so that the full axis region is
% covered.
svgImage(s, 'water_stones.jpg', 'xMidYMidslice', 'pic');
% Create a composite between bars and picture by union.
svgComposite(s, 'pic', 'SourceGraphic', 'atop', 'obj');
% To create the shadow we use the bars and use a blur filter
svgGaussianBlur(s, 'SourceAlpha', 5, 'blur');
% The blurred bars are now shifted by [10 10].
svgOffset(s, 'blur', [10 10], 'shade');
% Combine the shadow and 'picture bars' by applying the later
on top of the

% shadow.
svgComposite(s, 'obj', 'shade', 'over', 'front');
% For the background we use the same picture but make it brighter
svgComposite(s, 'pic', 'pic', 'arithmetic', 'background', [0 0.2 0 0]);
% As last step, we combine the foreground and background
svgComposite(s, 'front', 'background', 'over', 'final');
% Beautify the tick mark labels by rotating them by 90 deg
setting.svg.XTickLabelAngle = -90;
set(gca, 'UserData', setting);
% Finally, we save the result into a SVG file

Plot Result of the Example

Pixelgraphics of the SVG Plot.

View the SVG plot