removing units from plot - matlab

I try to remove the Matlab-given units from this plot but I don't find a way:
figure(1)
hold on
set(gcf,'PaperUnits','centimeters',...
'PaperSize',[15 9],...
'PaperPosition',[0 0 15 9]);
pzmap(LB); sgrid; grid on; axis equal;
title('');
xlabel('\sigma [rad/s]')
ylabel('\omega [rad/s]')
hold off
After that commands the xlabel looks like this: \sigma [rad/s] (seconds^-1). The seconds comes with pzmap. How can I remove them?
I found, some strange behavour:
If generate code by the figure plot manager I get this:
% Create xlabel
xlabel('\sigma [rad/s] (seconds^{-1})','Units','pixels');
Why???
Now I get it - without pzmap/pzplot
pol = pole(sys)
figure(1)
plot(real(pol(:)),imag(pol(:)),'x')
title('');
xlabel('\sigma [rad/s]');
ylabel('\omega [rad/s]');
sgrid

pzmap is a high-level convenience function, but it's not the best choice for this (it's also stored in a folder of obsolete functions in R2013a, so it may get marked for official removal in the future). Instead, let's create an example plot using pzplot directly instead of pzmap. This is still a plot function that does a lot under the hood, but it returns a handle, h, to the plot:
sys = rss(3,2,2);
h = pzplot(sys);
sgrid;
axis equal;
We can via the options of a pzplot with getoptions:
p = getoptions(h)
To set the labels and units as you desire, you might try this, using setoptions:
p.Title.String = '';
p.XLabel.String = '\sigma';
p.YLabel.String = '\omega';
setoptions(h,p);
I believe that the units of 'seconds-1' that the plot displays is equivalent to the 'rad/s' that you want to specify. I know that the two look is very different (I prefer being specific about radians myself), but that's a disadvantage of using such a plot function that tries to do everything for you. If you wanted to remove the default string or add another option, you'd likely have to do some low level hacking. An easier way around, might be to use the "Generate Code..." command ("Generate M-File..." in older versions") under the "File" menu in the figure's toolbar and edit the plot labels there (there's also a programmatic option for this on the File Exchange). Or you could output to postscript and edit that.
Alternatively, you can use pzoptions to create a list of options to pass to pzplot or pzmap (undocumented in the latter case):
p = pzoptions;
p.Title.String = '';
p.XLabel.String = '\sigma';
p.YLabel.String = '\omega';
sys = rss(3,2,2);
pzplot(sys,p);
sgrid;
axis equal;
You'll see that that for some reason the text size is much smaller in this case. pzplot and pzmap must set the font size to 10 themselves. You could easily do this.
Fore more on customizing this and related Control toolbox plots, see this article.

After intense low-level digging, there is actually a pretty simple way to override the default behavior.
p = pzplot(sys);
p.AxesGrid.XUnits = 'rad/s';
p.AxesGrid.YUnits = 'rad/s';
Changes appear to take effect immediately. I have even tried setting the value to nothing, i.e.
p.AxesGrid.XUnits = '';
and it effectively removes the annoying parenthesis with the units. Technically, matlab creates a custom-class element they store under the name AxesGrid in the resppack.mpzplot class instance, with some standard LTI-behavior. You can probably work around some stuff by "injecting" a script with the same name as one of the standard library functions, so that it will be called instead, and change things in there, but this is the closest I have come to removing those annoying units in a few lines.
As a side info, the AxesGrid object is initialized in
...\controllib\graphics\#resppack\#pzplot\initialize.m
should you want to check it out.

Related

inconsistent behavior with "undocumented Matlab" figure handles

I read the blog post on assigning transparency to plot markers. I tried the code on a simple example and all was well. Then I tried a tight loop, plotting a single point at a time (doing this to assign a different color to each point in the graph), and invariably within a few loop cycles, when I grab the "plothandle.MarkerHandle", it's empty. In these cases, the class of this empty object is Matlab.graphics.GraphicsPlaceholder
while when the operation is successful, the class is:
matlab.graphics.primitive.world.Marker
The basic loop follows. colormatrix assigns a [r,g,b] color to each data point.
hold on
opacity = 0.5;
for jk = 1:numel(idx
tmph = plot(foox(jk),fooy(jk),'o','color',colormatrix(jk,:) );
tmpk = tmph.MarkerHandle;
tmpk.FaceColorData = uint8(double(tmpk.EdgeColorData).* [1,1,1,opacity]');
tmpk.EdgeColorData = uint8(double(tmpk.EdgeColorData).* [1,1,1,opacity]');
end
I've tried things like clearing variables every loop, putting in a delay timer, and so on, with no luck. I'm using Matlab R2015a.
EDIT: here's a simple example. What I seem to be finding is that if I run the entire script, it always fails. If I break it into two pieces where noted and execute the second section with a separate key stroke (ctrl-enter or selectall/F9 in the IDE editor), everything works. And yes, I'm aware that "undocumented features" are risky , but since MathWorks still hasn't figured out that allowing transparency -- and indexed color assignments -- are good things for the plot function, I'm still looking for a better workaround than using patch to draw each data point.
figure
xfoo = 1:10;
yfoo = 2*xfoo;
tmph = plot(xfoo,yfoo,'p','color',[1,0,1]);
hold on
opacity = 0.7;
% wait a while here.
tmpk = tmph.MarkerHandle;
tmpk.FaceColorData = uint8(double(tmpk.EdgeColorData).*[1,1,1,opacity]');
tmpk.EdgeColorData = uint8(double(tmpk.EdgeColorData).*[1,1,1,opacity]');
The fact that the script seems to work if you wait a bit between the plot and the retrieval of tmph.MarkerHandle suggests that you have the same issue that was reported on the blog by a user running R2014b. Yair suggested to call drawnow after the plot:
figure
xfoo = 1:10;
yfoo = 2*xfoo;
tmph = plot(xfoo,yfoo,'p','color',[1,0,1]);
hold on
opacity = 0.7;
drawnow;
tmpk = tmph.MarkerHandle;
tmpk.FaceColorData = uint8(double(tmpk.EdgeColorData).*[1,1,1,opacity]');
tmpk.EdgeColorData = uint8(double(tmpk.EdgeColorData).*[1,1,1,opacity]');
The workaround didn't work for a user running R2015a, which doesn't sound promising, but the fact that waiting seems to help for you, is encouraging.

Alter tick labels to a '10^ - style' appearance

Though there are quite a lot of answered questions to that kind of issue, I wasn’t able to find a proper solution for my exact problem. Anyway.
I try to format my tick lables like shown in the following example:
I already found out that the ‘sprintf’ command is offering a possibility to alter the tick format. The closest I came to what I want is the 'e-notation' triggered by the following command:
set(ax,'YTickLabel',sprintf('%2.0e|',yticks))
However, I’d like my labels to appear just as shown in the example picture. Is there a simple way to do that?
Thank you very much in advance,
Joe
You could use Latex formatting and sprintfc to produce what you want. (You might not need sprintfc at all actually but that's a nice way of creating a cell array of strings with numbers in a single line.):
set(ax,'YTickLabels',sprintfc('10^{%i}',yticks)
In a general example (here with x-axis formatted):
clear
clc
close all
x = 0:100000;
y = log(x);
figure
semilogx(x,y)
xt = get(gca,'XTick');
set(gca,'XTickLabels',sprintfc('10^{%i}',0:numel(xt)-1))
outputs the following:
Am I missing something? Why not use semilogy?
x = -3:0;
y = 10.^x;
semilogy(x, y);
set(gca, 'YMinorGrid', 'on')

Change markers in dynamic plots

What I have:
hold on
for i =1:length(tspan)
var{i} = Tv(:,i);
str{i} = ['t = ',num2str(tspan(i)), ' s'];
plot(z,var{i},'DisplayName',str{i});
end
legend('-DynamicLegend');
This works perfectly (thanks to this), but it prints out all blue lines.
I tried to set up a colormap (the default) and use it like this, but the output was the same
plot(z,var{i},'DisplayName',str{i},'Color', colormap(i,:));
And I would also like to see different markers for each plot. How is it possible to change them?
EDIT
Thanks to ironzionlion I fixed the colors. How can I do the same with markers?
According to the post that you mention, you have to define that you will need i different colours in your plot. This can be done by using colors = hsv(i)
Your plot sentence will then be: plot(z,var{i},'DisplayName',str{i},'Color', colors(i,:));
Update
I am not aware of the existance of "markermap". You could fix the problem just by define upfront the different markers that you want (quick and dirty solution): mrk={'o','+','*','.'};
Then you will plot by selecting each time the corresponding marker:
plot(z,var{i},'DisplayName',str{i},'Color', cmap(i,:),'Marker', mrk{i});

Location based segmentation of objects in an image (in Matlab)

I've been working on an image segmentation problem and can't seem to get a good idea for my most recent problem.
This is what I have at the moment:
Click here for image. (This is only a generic example.)
Is there a robust algorithm that can automatically discard the right square as not belonging to the group of the other four squares (that I know should always be stacked more or less on top of each other) ?
It can sometimes be the case, that one of the stacked boxes is not found, so there's a gap or that the bogus box is on the left side.
Your input is greatly appreciated.
If you have a way of producing BW images like your example:
s = regionprops(BW, 'centroid');
centroids = cat(1, s.Centroid);
xpos = centroids(:,1); should then be the x-positions of the boxes.
From here you have multiple ways to go, depending on whether you always have just one separated box and one set of grouped boxes or not. For the "one bogus box far away, rest closely grouped" case (away from Matlab, so this is unchecked) you could even do something as simple as:
d = abs(xpos-median(xpos));
bogusbox = centroids(d==max(d),:);
imshow(BW);
hold on;
plot(bogusbox(1),bogusbox(2),'r*');
Making something that's robust for your actual use case which I am assuming doesn't consist of neat boxes is another matter; as suggested in comments, you need some idea of how close together the positioning of your good boxes is, and how separate the bogus box(es) will be.
For example, you could use other regionprops measurements such as 'BoundingBox' or 'Extrema' and define some sort of measurement of how much the boxes overlap in x relative to each other, then group using that (this could be made to work even if you have multiple stacks in an image).

Change label of data tips on bode nichols diagram

When we plot a bode/nichols locus, the name of workspace variable is used
tmp=ss(1,1,1,0);
nichols(tmp);
will use 'tmp' as label.
When using more complex data, matlab is using 'untitled1','untitled2',...
tmp={ss(1,1,1,0) , ss(1.2,1,1,0)};
nichols(tmp{:});
How can I change this label programmatically?
Ideally, I'd like a solution working with Matlab 6.5.1, but I'm also interested in solutions restricted to newer versions.
You can modify the labels programmatically via their graphics handles. It looks like the values you want to change are the DisplayName property of some of the children of the current axis. So in your first example, I can change the display name like this:
ch = get(gca,'Children');
set(ch(1),'DisplayName','Fred');
In general, I'm not sure how to predict which children of the current axis are the ones you need to change. For the second example you give, the two curves appear to be the second and third children when I run your code.