GNU Octave Matlab: Plot tick labeling - matlab

I am making a frequency plot and I would like some help on tick labeling.
Here is what I have:
semilogx([200,1000,5000], [0,6,0]);
xlim([20 20000]);
sc = [20:10:100,125:25:175];
scale = [sc,sc*10,sc*100, 20000];
xticks(scale);
xticklabels(scale);
set(gca,'XMinorTick','Off')
grid on;
set (gca, "xminorgrid", "off")
xlabel('frequency (Hz)');
ylabel('dB');
How can I make all numbers from 1000 and upwards appear as 1K, 2K, 5K and so on?
How could I make the lines on 50,100,200,500,1K,2K,5K,10K appear thicker/more black?

Octave approach (probably works on matlab too though)
I wouldn't rely on latex trickery to do this to be honest.
Here is the way I usually do stuff like this.
Effectively, because the axis labels object is considered a single object, and you cannot split it into parts, the trick is to overlay an invisible, bare-minimum axes object defining only the labels you want, and treat those as you'd like (e.g. adjust its fontweight, fontsize, xcolor, etc etc).
H = semilogx([200,1000,5000], [0,6,0]);
A = gca();
B = axes();
subscale = [20:10:100,125:25:175];
scale = [subscale,subscale * 10,subscale * 100, 20000];
ScaleTextLabels = {};
for i = 1 : length( scale )
if scale(i) >= 1000, ScaleTextLabels{i} = sprintf("%dk", scale(i) / 1000 );
else, ScaleTextLabels{i} = num2str( scale(i) );
end
end
SpecialTickLabels = { '50', '100', '200', '500', '1k', '2k', '5k', '10k'};
ScaleIndices = 1 : length( ScaleTextLabels );
SpecialIndices = nthargout( 2, #ismember, SpecialTickLabels, ScaleTextLabels );
NormalIndices = setdiff( ScaleIndices, SpecialIndices );
set( A, 'xgrid', 'on', 'xlabel', 'frequency (Hz)', 'xlim', [20 20000] , 'xminorgrid', 'off', 'xminortick', 'off', 'xticklabel', ScaleTextLabels(NormalIndices), 'xtick', scale(NormalIndices) , 'ylabel', 'dB', 'gridlinestyle', ':', 'gridcolor', 'k', 'gridalpha', 0.5 );
set( B, 'xgrid', 'on', 'xlabel', '' , 'xlim', get( A, 'xlim' ), 'xminorgrid', 'off', 'xminortick', 'off', 'xticklabel', ScaleTextLabels(SpecialIndices), 'xtick', scale(SpecialIndices), 'ylabel', '' , 'color', 'none', 'fontsize', 12, 'fontweight', 'bold', 'position', get( A, 'position'), 'xcolor', [0,0,0], 'xscale', 'log', 'ylim', get( A, 'ylim'), 'ytick', [], 'gridlinestyle', '--', 'gridcolor', 'k', 'gridalpha', 0.8 );
This "layers of transparent axes objects" technique is very useful to keep in mind in general, it allows great flexibility when designing complex graphs. :)

In MATLAB
*I unfortunately could not yet find how to bold the specific lines
Adding the following code allows the ticks to converted to the new names/format suggested in part 1. For part 2 the best I could find out right now is bolding the specific numbers, unfortunately not the specific ticks/lines. Here \bf indicates which labels are to be bolded. All the names will correspond to the positions set originally by your axis vector scale. The last line in the code below indicates the replacement of the current axis, gca.
semilogx([200,1000,5000],[0,6,0]);
sc = [20:10:100,125:25:175];
scale = [sc,sc*10,sc*100, 20000];
Current_Axis = gca;
Current_Axis.XMinorTick = 'off';
xlabel('frequency (Hz)'); ylabel('dB');
xlim([20 20000]);
grid on;
X_Scale_Names = {'\bf20'; '30'; '40'; '\bf50'; '60';
'70';'80';'90';'\bf100';'125';'150';'175';'\bf200';'300';'400';
'500';'600';'700';'800';'900';'\bf1K';'1.25K';'1.5K';'1.75K';
'\bf2K';'3K';'4K';'\bf5K';'6K';'7K';'8K';'9K';'\bf10K';'12.5K';'15K';
'17.5K';'20K'};
To Adjust More Grid and Axis Properties:
Current_Axis = gca;
set(Current_Axis,'xtick',scale,'xticklabel',X_Scale_Names);
Current_Axis.LineWidth = 1;
Current_Axis.GridColor = 'k';
Current_Axis.GridAlpha = 0.5;
Ran using MATLAB R2019b

I did it like this:
semilogx([200,1000,5000], [0,6,0]);
xlim([20 20000]);
sc = [20:5:35,40:10:100,125:25:175];
scale = [sc,sc*10,sc*100, 20000];
xticks(scale);
xticklabels(scale);
set(gca,'XMinorTick','Off')
grid on;
set(gca,'gridlinestyle',':');
set(gca,'gridalpha',0.6);
set (gca, "xminorgrid", "off");
xg = [50,100,200,500,1000,2000,5000,10000]; #highlight grids
xx = reshape([xg;xg;NaN(1,length(xg))],1,length(xg)*3);
yy = repmat([ylim() NaN],1,length(xg));
line(xx,yy,'Color',[0.65,0.65,0.65]);
xlabel('frequency (Hz)');
ylabel('dB');
X_Scale_Names = {'\fontsize{11}\bf20'; '25'; '30';'35';'40'; '\fontsize{11}\bf50'; '60';
'70';'80';'90';'\fontsize{11}\bf100';'125';'150';'175';'\fontsize{11}\bf200';'250';'300';'350';'400';
'\fontsize{11}\bf500';'600';'700';'800';'900';'\fontsize{11}\bf1K';'1.25K';'1.5K';'1.75K';
'\fontsize{11}\bf2K';'2.5K';'3K';'3.5K';'4K';'\fontsize{11}\bf5K';'6K';'7K';'8K';'9K';'\fontsize{11}\bf10K';'12.5K';'15K';
'17.5K';'\fontsize{11}\bf20K'};
set(gca,'xtick',scale,'xticklabel',X_Scale_Names);
But I don't think this is the best/fastest/easiest way to do it...

Related

Smoothing/interpolation of a 3D surface colormap

I'm trying to smooth or interpolate away the "steps" building up to a intensity maximum, and at the same time preserving the shape of the surface so that the estimated width does not get shifted or changed. I have tried interpolating to a higher resolution using 2D spline, but the "steps" are still present and the estimated maxima region changes. Could someone please help me in smoothing/interpolating away those steps to make the surface look more continuous while conserving the surface shape?
As always, many thanks in advance for any advice or help!!
Uninterpolated
Interpolated
Attempt
conversion = 1/49.0196; % in mm/pixel
new_xData = linspace(xData(1), xData(end), 10*length(xData));
new_yData = linspace(yData(1), yData(end), 10*length(yData)).';
new_zData = interp2(xData, yData, zData, new_xData, new_yData, 'spline');
xData = new_xData; yData = new_yData; zData = new_zData;
h = figure();
surf(xData, yData, zData, 'LineStyle', 'none', 'FaceColor', 'interp');
xlim([xData(1) xData(end)])
ylim([yData(1) yData(end)])
zlim([0 ceil(max(zData(:)))]);
colormap(jet(4096))
xlabel('mm'); ylabel('mm'); zlabel('ln(Intensity)');
set(gca,'Ydir', 'reverse')
grid on
shading interp
view(3); axis vis3d; camlight;
% Change x and y tick labels from pixels to mm
addMM = #(x) sprintf('%.1f', x * conversion);
xticklabels(cellfun(addMM,num2cell(xticks'),'UniformOutput',false));
yticklabels(cellfun(addMM,num2cell(yticks'),'UniformOutput',false));
% Find maxium intensity region and plot it
hold on
maxValue = max(zData(:));
[rowsOfMaxes, colsOfMaxes] = find(zData == maxValue);
x_minind = min(colsOfMaxes(:));
x_maxind = max(colsOfMaxes(:));
p = zeros(1, 2);
p(1) = plot3(linspace(xData(x_minind), xData(x_minind), ...
length(yData)), yData, zData(:, x_minind), 'Linestyle', '-.', ...
'Color', 'black', 'LineWidth', 2.0);
p(2) = plot3(linspace(xData(x_maxind), xData(x_maxind), ...
length(yData)), yData, zData(:, x_maxind), 'Linestyle', '-.', ...
'Color', 'black', 'LineWidth', 2.0);
hold off
legend(p(1), 'Estimated width of maximum intensity', 'Location', 'North')
set(gca, 'FontSize', 14)

Multiple y axes in MATLAB (axis alignment issue when printing)

Here is a sample of a strange problem. I'd like to plot curves with multiple y axes and I use a fairly common method in MATLAB.
function plot_test()
clear;
savefile = 1;
scrsz = get(0,'ScreenSize');
figure('Position',[1 1 0.9 * scrsz(3) 0.9 * scrsz(4)]);
hold on;
box on;
x = 0:1:10;
h1 = plot(x, x.^2 , 'r', 'LineWidth', 2.5);
%Axis label
xlabel('XLabel', 'FontSize', 20, 'Interpreter', 'latex');
ylabel('YLabel', 'FontSize', 20, 'Interpreter', 'latex');
set(gca, 'FontSize', 20, 'LineWidth', 3);
ax1 = gca;
ax2 = axes('Position',get(ax1,'Position'),'XAxisLocation','top','YAxisLocation','right','Color','none','XColor','none','YColor','k');
linkaxes([ax1 ax2],'x');
hold on
box on;
h2 = plot(x, x, 'b', 'Parent', ax2, 'LineWidth', 2.5);
ylabel('Second YLabel', 'FontSize', 20, 'Interpreter', 'latex');
set(gca, 'FontSize', 20, 'LineWidth', 3);
hl=legend([h1 h2],{'First Line','Second Line'});
set(hl,'FontSize',15,'Location','Northwest', 'Orientation','Vertical')
%Save pdf
if savefile
% Backup previous settings
prePaperType = get(gcf,'PaperType');
prePaperUnits = get(gcf,'PaperUnits');
preUnits = get(gcf,'Units');
prePaperPosition = get(gcf,'PaperPosition');
prePaperSize = get(gcf,'PaperSize');
% Make changing paper type possible
set(gcf,'PaperType','<custom>');
% Set units to all be the same
set(gcf,'PaperUnits','inches');
set(gcf,'Units','inches');
% Save the pdf
print -dpdf Test.pdf;
% Restore the previous settings
set(gcf,'PaperType',prePaperType);
set(gcf,'PaperUnits',prePaperUnits);
set(gcf,'Units',preUnits);
set(gcf,'PaperPosition',prePaperPosition);
set(gcf,'PaperSize',prePaperSize);
end
The objective is to print a PDF of the figure and save it in the same folder as Test.pdf. This is accomplished but the axes are misaligned. On my Windows machine it looks horrible while on a Mac it looks almost okay (but if you look closely, the y-axes are indeed misaligned at the bottom).
This only happens when I use a second axis. Without that, all this runs perfectly. Any idea why?
Okay, so I found a way: The trick is to use plotyy. Sample code below
function plot_test2()
clear;
savefile = 1;
scrsz = get(0,'ScreenSize');
figure('Position',[1 1 0.9 * scrsz(3) 0.9 * scrsz(4)]);
hold on;
box on;
x=(0:1:10);
y1=x;
y2=x.^2;
[hAx, hLine1, hLine2] = plotyy(x,y1,x,y2);
%Axis label
xlabel(hAx(1),'XLabel', 'FontSize', 20, 'Interpreter', 'latex', 'Color', 'k');
ylabel(hAx(1),'YLabel', 'FontSize', 20, 'Interpreter', 'latex', 'Color', 'k');
ylabel(hAx(2),'Second YLabel', 'FontSize', 20, 'Interpreter', 'latex');
set(hAx,{'ycolor'},{'k';'k'})
set(hAx,{'FontSize'},{20;20}, {'LineWidth'}, {3;3})
set(hLine1,'LineWidth', 3)
set(hLine2,'LineWidth', 3)
set(hLine1,'Color', 'r')
set(hLine2,'Color', 'b')
hl=legend([hLine1 hLine2],{'First Line','Second Line'});
set(hl,'FontSize',15,'Location','Northwest', 'Orientation','Vertical')
%Save pdf
if savefile
% Backup previous settings
prePaperType = get(gcf,'PaperType');
prePaperUnits = get(gcf,'PaperUnits');
preUnits = get(gcf,'Units');
prePaperPosition = get(gcf,'PaperPosition');
prePaperSize = get(gcf,'PaperSize');
% Make changing paper type possible
set(gcf,'PaperType','<custom>');
% Set units to all be the same
set(gcf,'PaperUnits','inches');
set(gcf,'Units','inches');
% Save the pdf
print -dpdf Test.pdf;
% Restore the previous settings
set(gcf,'PaperType',prePaperType);
set(gcf,'PaperUnits',prePaperUnits);
set(gcf,'Units',preUnits);
set(gcf,'PaperPosition',prePaperPosition);
set(gcf,'PaperSize',prePaperSize);
end

Matlab - remove xtick with a bar plot

I get the following figure with a bar plot :
I would like to remove the XTick with red circles but keep Xtick for the middle of each group of bars (i.e 1024, 10240, 102400, 1024000, 10240000).
I generate this image with the Matlab script below :
x = load('performances.txt');
% Get Runtimes
for i = 1:6
time_seq(1:5,i) = x((i-1)*5+1:i*5,3);
time_gpu(1:5,i) = x((i-1)*5+1:i*5,4);
speedup(1:5,i) = time_seq(1:5,i)./time_gpu(1:5,i);
end
% X axis
sizeArray = [1024 10240 102400 1024000 10240000 102400000]
figure(1);
% Get Histogram
h = bar(log10(sizeArray),log10(speedup(1:5,:)')); % get histogram
% Log10 for x-axis and xtick
set(gca,'Xtick',log10(1024):1:log10(1.024*10^8))
set(gca,'Xticklabel',10.^get(gca,'Xtick'));
set(h(1),'facecolor',[0.5 0.5 1]);
set(h(2),'facecolor',[1 0.5 0.5]);
set(h(3),'facecolor',[0.5 1 0.5]);
set(h(4),'facecolor',[0.5 0.5 0.5]);
set(h(5),'facecolor',[1 0.5 1]);
hPatch = findobj(h,'Type','patch');
set(hPatch,'facealpha',1);
grid on;
title('Benchmark GPU vs CPU');
% Size of WorkGroup
h = legend('N=16','N=32','N=64','N=128','N=256');
v = get(h,'title');
set(v,'string','WorkGroup size');
% Place legend
rect = [0.6,0.25,0.2,0.2];
set(h,'Position',rect,'color','w');
hPatch = findobj(h,'Type','patch');
set(hPatch,'facealpha',1);
xlabel('log(Array size)');
ylabel('log(Speedup)');
% Make right y-axis visible
ax1 = gca;
ax2 = axes('Position', get(ax1, 'Position'));
set(ax2, 'YAxisLocation', 'right', 'Color', 'none', 'XTickLabel', []);
set(ax2, 'YLim', get(ax1, 'YLim'));
I tried different things but couldn't make them disappear, anyone would have an idea or a clue ?
Thanks
It happens because of the last 3 lines in your code where you add the Y-axis from the right side. You need to add the following line at the end:
set(ax2,'XTick',[]);
Here is my result (with fake data):

smoothing graph line in Matlab

I have the following graph and i would like to make it more pleasing to the eyes by smoothing the graph. is it possible ?
tempyr = 1880:1:2014;
temperature = temp(1:2, 1:135);
Tempval = {'Annual Mean','5 Year Mean'}
TH = zeros(size(Tempval));
hold on
TH = plot( tempyr', temperature', '-o', 'Marker', '.');
xlabel( 'year', 'fontsize', 24); ylabel( 'Temperature Anomaly (Degree Cel)', 'fontsize', 24 );
legend(TH, Tempval)
grid on
Ideal graph.
Try
TH = plot( tempyr', temperature', '-o', 'Marker', '.','LineSmoothing','on');
and also have a look here, especially the export_fig reference might prove useful.

Matlab: Problem with ticks when setting minor grid style and two y-axis

with help of the community in this thread: Minor grid with solid lines & grey-color
I got it to work to set minor grid lines as solid and coloured style. But when adding a second y-axes it just messes up the y-ticks on the right axis! heres the example code:
x = linspace(0, 10, 11);
y1 = x.^3+1;
y2 = x+1;
y3 = y1./y2+5;
% plotte: http://www.mathworks.com/help/techdoc/ref/linespec.html
myfig = figure('Position', [500 500 445 356]); %[left, bottom, width, height]:
ax1 = axes('Position',[0.13 0.18 0.75 0.75]);
hold on
p1 = plot(x,y1,'x--r');
p2 = plot(x,y2,'*-b');
xlim([0 max(x)]);
ylim([0 max([max(y1) max(y2)])]);
col=.85*[1 1 1];
%# create a second transparent axis, same position/extents, same ticks and labels
ax2 = axes('Position',get(ax1,'Position'), ...
'Color','none', 'Box','on', ...
'XTickLabel',get(ax1,'XTickLabel'), 'YTickLabel',get(ax1,'YTickLabel'), ...
'XTick',get(ax1,'XTick'), 'YTick',get(ax1,'YTick'), ...
'XLim',get(ax1,'XLim'), 'YLim',get(ax1,'YLim'));
%# show grid-lines of first axis, give them desired color, but hide text labels
set(ax1, 'XColor',col, 'YColor',col, ...
'XMinorGrid','on', 'YMinorGrid','on', ...
'MinorGridLineStyle','-', ...
'XTickLabel',[], 'YTickLabel',[]);
%# link the two axes to share the same limits on pan/zoom
linkaxes([ax1 ax2],'xy');
ax3 = axes('Position',get(ax1,'Position'),...
'XAxisLocation','top',...
'YAxisLocation','right',...
'Color','none',...
'XTickLabel', [],...
'XColor','k','YColor','k');
%# link the two axes to share the same limits on pan/zoom
linkaxes([ax1 ax2 ax3], 'x');
ylabel(ax3, 'Speedup []');
ylim(ax3, [0 max(y3)]);
hold on
p3 = plot(x,y3,'s-.m','Parent',ax3);
hleg = legend([p1 p2 p3], {'CPU', 'GPU', 'Speedup'}, 'Location', 'NorthWest');
xlabel(ax2, 'N_{Funcs}');
ylabel(ax2, 't [s]');
set(hleg, 'FontAngle', 'italic')
and how it looks like:
Its simpler than you think: when you create the second axis ax2, set the 'Box' property to 'off' instead of 'on'.
Even more, you can simplify that part and create it as:
ax2 = copyobj(ax1,myfig);
delete( get(ax2,'Children') )
set(ax2, 'Color','none', 'Box','off')
The 2nd y-axis is "messed up" because the automatically generated YTick from y3 does not agree with the YTick from y1 and y2.
If this view is final (meaning you don't have to zoom in/zoom out or move the plot), you can manually define the YTick of ax3 to match those of ax1
ax3 = axes('Position',get(ax1,'Position'),...
'XAxisLocation','top',...
'YAxisLocation','right',...
'Color','none',...
'XTickLabel', [],...
'YTick', [0:max(y3)/5:max(y3)], ... %% Define 6 YTick (including 0) like ax1
'XColor','k','YColor','k');