Related
I want to draw a contour plot of a plane and a surface with legends. Plotting two surfaces in the same figure create the same legends. I want to change the resulting ellipses in the legend. Can I draw parallel lines instead of ellipses on the legend chart?
This is a sample source code:
[X,Y] = meshgrid(-3:.1:3);
Z1 = peaks(X,Y);
Z2 = 2*X+3*Y+4;
contour(X,Y,Z1)
colormap jet
shading interp
axis([-3 3 -3 3])
hold on
contour(X,Y,Z2)
legend('surface','plane')
Honestly there's no good way to get an image in front.
There are a few work arounds that I could show you, but you'll have to be specific.
The easiest things are, getting a colorbar or a rectangular box (1 color) infront of text. These are standard options for datatypes/legend entries.
I suggest you start off with giving your surfaces/countours handles.
h1=surf(....);
h2=plot(....);
lgd=legend([h1, h2, ....],[entries']);
For text with white in front you can do
h_separator1 = plot(NaN,NaN,'.','Color',[1 1 1]);
lgd=legend([h_separator1],['text']);
Here's an example you can run that let's you do the things I just said
clear all; % just for ease of quickly plotting this
close all; %closing all figures
myc=[1 1 1; 0 0 0; 1 1 1; 0 0 0; 1 1 1]; %this is want we will use to draw the paralel lines, can be of any color, just replace the zeros, with the respective values of the colors you want, 0 0 0 is black
x = [0 0 0 0]; %making it have 0 area and thus invisible
y = [0 0 0 0];
c = [0 0.33 0.66 1]; %lets you add a colorbar
figure
colormap(myc); %update the figure to use your colormap
hold on
h3 = plot(NaN,NaN,'Color','none'); %blank entry
h4 = plot(NaN,NaN,':'); % entry with dotted line, color "pseudo random"
h1=patch(x,y,'red','EdgeColor','none'); %For a rectangular color entry in legend
h2=patch(x,y,c,'EdgeColor','none'); %lets you add the colorbar, later use to place inside the legend as paralel lines
[lgd,OBJH,OUTH,OUTM]=legend([h1,h3,h2,h4],{'HI your text here','Nothing','paralel lines','line'}); %the lgd handle let's you later modify properties of the legend
hcb=colorbar; %the colorbar can still be modified, to have no number or a different scale, color, etc.
hcm=OBJH(5)
xlim([0 1])
ylim([0 1])
lpos=lgd.Position; % get position legend
lnr=numel(OUTH); %legend entries is useful
lhstp=lpos(4)/(lnr+1); %heigth step
hax=gca;
axpos=hax.Position; %to shift position because the colorbar is placed in the figure and not within the axes in comparison to the legend
% placing at by example 3rd entry
wdentry=0.04; %at the moment trial and error depends on width on legend box which is based on amount of characters and text size.
p1=axpos(1)+lpos(1)+0.01;
p2=lpos(2)+3/2*lhstp;
p3=wdentry;
p4=lhstp-0.01;
hcb.TickLabels=[]; %remove tick labels
hcb.Ticks=[]; %remove ticks
hcb.Color=[1 1 1]; %white border around it to make it "semi-invisible"
hcb.Position=[p1 p2 p3 p4];
I'm trying to make a contour that follows the edges of the 'pixels' in a pcolor plot in Matlab. This is probably best explained in pictures. Here is a plot of my data. There is a distinct boundary between the yellow data (data==1) and the blue data (data==0):
Note that this is a pcolor plot so each 'square' is essentially a pixel. I want to return a contour that follows the faces of the yellow data pixels, not just the edge of the yellow data.
So the output contour (green line) passes through the mid-points of the face (red dots) of the pixels.
Note that I don't want the contour to follow the centre points of the data (black dots), which would do something like this green line. This could be achieved easily with contour.
Also, if it's any help, I have a few grids which may be useful. I have the points in the middle of the pixels (obviously, as that's what I've plotted here), I also have the points on the corners, AND I have the points on the west/east faces and the north/south faces. IF you're familiar with Arakawa grids, this is an Arakawa-C grid, so I have the rho-, u-, v- and psi- points.
I've tried interpolation, interweaving grids, and a few other things but I'm not having any luck. Any help would be HUGELY appreciated and would stop me going crazy.
Cheers, Dave
EDIT:
Sorry, I simplified the images to make what I was trying to explain more obvious, but here is a larger (zoomed out) image of the region I'm trying to separate:
As you can see, it's a complex outline which heads in a "southwest" direction before wrapping around and moving back "northeast". And here is the red line that I'd like to draw, through the black points:
You can solve this with a couple of modifications to a solution I posted to a related question. I used a section of the sample image mask in the question for data. First, you will need to fill the holes in the mask, which you can do using imfill from the the Image Processing Toolbox:
x = 1:15; % X coordinates for pixels
y = 1:17; % Y coordinates for pixels
mask = imfill(data, 'holes');
Next, apply the method from my other answer to compute an ordered set of outline coordinates (positioned on the pixel corners):
% Create raw triangulation data:
[cx, cy] = meshgrid(x, y);
xTri = bsxfun(#plus, [0; 1; 1; 0], cx(mask).');
yTri = bsxfun(#plus, [0; 0; 1; 1], cy(mask).');
V = [xTri(:) yTri(:)];
F = reshape(bsxfun(#plus, [1; 2; 3; 1; 3; 4], 0:4:(4*nnz(mask)-4)), 3, []).';
% Trim triangulation data:
[V, ~, Vindex] = unique(V, 'rows');
V = V-0.5;
F = Vindex(F);
% Create triangulation and find free edge coordinates:
TR = triangulation(F, V);
freeEdges = freeBoundary(TR).';
xOutline = V(freeEdges(1, [1:end 1]), 1); % Ordered edge x coordinates
yOutline = V(freeEdges(1, [1:end 1]), 2); % Ordered edge y coordinates
Finally, you can get the desired coordinates at the centers of the pixel edges like so:
ex = xOutline(1:(end-1))+diff(xOutline)./2;
ey = yOutline(1:(end-1))+diff(yOutline)./2;
And here's a plot showing the results:
imagesc(x, y, data);
axis equal
set(gca, 'XLim', [0.5 0.5+size(mask, 2)], 'YLim', [0.5 0.5+size(mask, 1)]);
hold on;
plot(ex([1:end 1]), ey([1:end 1]), 'r', 'LineWidth', 2);
plot(ex, ey, 'k.', 'LineWidth', 2);
Take a look at the following code:
% plotting some data:
data = [0 0 0 0 0 0 1 1
0 0 0 0 0 1 1 1
0 0 0 0 1 1 1 1
0 0 0 0 0 1 1 1
0 0 0 0 1 1 1 1
0 0 0 0 1 1 1 1
0 0 0 0 1 1 1 1];
p = pcolor(data);
axis ij
% compute the contour
x = size(data,2)-cumsum(data,2)+1;
x = x(:,end);
y = (1:size(data,1));
% compute the edges shift
Y = get(gca,'YTick');
y_shift = (Y(2)-Y(1))/2;
% plot it:
hold on
plot(x,y+y_shift,'g','LineWidth',3,'Marker','o',...
'MarkerFaceColor','r','MarkerEdgeColor','none')
It produces this:
Is this what you look for?
The most important lines above is:
x = size(data,2)-cumsum(data,2)+1;
x = x(:,end);
which finds the place of shifting between 0 to 1 for every row (assuming there is only one in a row).
Then, within the plot I shift y by half of the distance between two adjacent y-axis tick, so they will be placed at the center of the edge.
EDIT:
After some trials with this kind of data, I have got this result:
imagesc(data);
axis ij
b = bwboundaries(data.','noholes');
x = b{1}(:,1);
y = b{1}(:,2);
X = reshape(bsxfun(#plus,x,[0 -0.5 0.5]),[],1);
Y = reshape(bsxfun(#plus,y,[0 0.5 -0.5]),[],1);
k = boundary(X,Y,1);
hold on
plot(X(k),Y(k),'g','LineWidth',3,'Marker','o',...
'MarkerFaceColor','r','MarkerEdgeColor','none')
It's not perfect, but may get you closer to what you want in a more simple approach:
OK, I think I've solved it... well close enough to be happy.
First I take the original data (which I call mask_rho and use this to make masks mask_u, mask_v, which is similar to mask_rho but is shifted slightly in the horizontal and vertical directions, respectively.
%make mask_u and mask_v
for i = 2:size(mask_rho,2)
for j = 1:size(mask_rho,1)
mask_u(j, i-1) = mask_rho(j, i) * mask_rho(j, i-1);
end
end
for i = 1:size(mask_rho,2)
for j = 2:size(mask_rho,1)
mask_v(j-1, i) = mask_rho(j, i) * mask_rho(j-1, i);
end
end
I then make modified masks mask_u1 and mask_v1 which are the same as mask_rho but averaged with the neighbouring points in the horizontal and vertical directions, respectively.
%make mask which is shifted E/W (u) and N/S (v)
mask_u1 = (mask_rho(1:end-1,:)+mask_rho(2:end,:))/2;
mask_v1 = (mask_rho(:,1:end-1)+mask_rho(:,2:end))/2;
Then I use the difference between the masks to locate places where the masks change from 0 to 1 and 1 to 0 in the horizontal direction (in the u mask) and in the vertical direction (in the v mask).
% mask_u-mask_u1 gives the NEXT row with a change from 0-1.
diff_mask_u=logical(mask_u-mask_u1);
lon_u_bnds=lon_u.*double(diff_mask_u);
lon_u_bnds(lon_u_bnds==0)=NaN;
lat_u_bnds=lat_u.*double(diff_mask_u);
lat_u_bnds(lat_u_bnds==0)=NaN;
lon_u_bnds(isnan(lon_u_bnds))=[];
lat_u_bnds(isnan(lat_u_bnds))=[];
%now same for changes in mask_v
diff_mask_v=logical(mask_v-mask_v1);
lon_v_bnds=lon_v.*double(diff_mask_v);
lon_v_bnds(lon_v_bnds==0)=NaN;
lat_v_bnds=lat_v.*double(diff_mask_v);
lat_v_bnds(lat_v_bnds==0)=NaN;
lon_v_bnds(isnan(lon_v_bnds))=[];
lat_v_bnds(isnan(lat_v_bnds))=[];
bnd_coords_cat = [lon_u_bnds,lon_v_bnds;lat_u_bnds,lat_v_bnds]'; %make into 2 cols, many rows
And the result grabs all the coordinates at the edges of the boundary:
Now my answer goes a bit awry. If I plot the above vector as points plot(bnd_coords_cat(:,1),bnd_coords_cat(:,2),'kx' I get the above image, which is fine. However, if I join the line, as in: plot(bnd_coords_cat(:,1),bnd_coords_cat(:,2),'-' then the line jumps around, as the points aren't sorted. When I do the sort (using sort and pdist2) to sort by closest points, Matlab sometimes chooses odd points... nevertheless I figured I'd include this code as an appendix, and optional extra. Someone may know a better way to sort the output vectorbnds_coords_cat:
% now attempt to sort
[~,I]=sort([lon_u_bnds,lon_v_bnds]);
bnd_coords_inc1 = bnd_coords_cat(I,1);
bnd_coords_inc2 = bnd_coords_cat(I,2);
bnd_coords = [bnd_coords_inc1,bnd_coords_inc2];
bnd_coords_dist = pdist2(bnd_coords,bnd_coords);
bnd_coords_sort = nan(1,size(bnd_coords,1));
bnd_coords_sort(1)=1;
for ii=2:size(bnd_coords,1)
bnd_coords_dist(:,bnd_coords_sort(ii-1)) = Inf; %don't go backwards?
[~,closest_idx] = min(bnd_coords_dist(bnd_coords_sort(ii-1),:));
bnd_coords_sort(ii)=closest_idx;
end
bnd_coords_final(:,1)=bnd_coords(bnd_coords_sort,1);
bnd_coords_final(:,2)=bnd_coords(bnd_coords_sort,2);
Note that the pdist2 method was suggested by a colleague and also from this SO answer, Sort coordinates points in matlab. This is the final result:
To be honest, plotting without the line is fine. So as far as I'm concerned this is close enough to be answered!
I'm plotting line charts with the x-axis representing the timeline. Some of the dataranges should be marked as unreliable. For showig this it would be great to change the background color over these x-ranges to some bringt yellow or red. Is there an option of doing that?
I know that could us gscatter instead of gplot for using colors. But then only the data is colored and not the background and I don't get linecharts but point charts. Colored data would be an acceptable alternative, but I need line charts.
It is always possible to make a filled polygon plot. Try,
a = 0:0.1:1;
b = a;
figure;
plot(a,b,'r');
hold on;
x = [0 0 1 1 0];
y = [0 1 1 0 0];
h = fill(x,y,'g');
set(h,'FaceAlpha',0.1); % or alpha(h,0.1); would also work
I need to create an nth-order Hadamard matrix, row double it, within each row randomly permute the elements of the matrix, and then display it. So far, I have accomplished all of these things. What I end up with when I imshow(matrix) is a nice picture of black and white boxes. But I haven't figured out how to insert a fine line to divide each row. I can create something like the first image on the left, but not the image on the right (these are Figures 1 and 2 from this paper)
Any help or comments would be thoroughly appreciated.
I've found using vector approaches (e.g., patch and rectangle) for this sort of problem unnecessarily challenging. I think that it's more straightforward to build a new image. This avoids floating-point rounding issues and other things that crop up with vector graphics. My solution below relies on some functions in the Image Processing Toolbox, but is simple and fast:
% Create data similarly to #TryHard
H = hadamard(48);
C = (1+[H;-H])/2;
rng(0); % Set seed
C(:) = C(randperm(numel(C))); % For demo, just permute all values, not rows
% Scale image and lines
scl = 10; % Amount to vertically scale each row
pad = 2; % Number of pixels to add between each row
C = imresize(C,scl,'nearest');
C = blockproc(C,[scl size(C,2)],#(x)[x.data;zeros(pad,size(C,2))]);
C = C(1:end-pad,:); % Remove last line added
% Dispay image
imshow(C)
This results in an image like this
The scl and pad parameters can be easily adjusted to obtain different sizes and relative sizes. You can call imresize(...,'nearest') again after adding the lines to further scale the image if desired. The blocproc line could potentially be made more efficient with various options (see the help). It could also be replaced by calls to im2col and col2im, which possibly could be faster, if messier.
I did not try the code, but I think that something like that should work:
sizeOfACube = 6;
numberOfRows = 47;
RGB = imread('image.png');
RGB = imresize(A, [(numRows+numberOfRows) numCols]);
for i=1:1:NumberOfRows
RGB(i*6,:,:) = 0;
end
imagesc(RGB);
imwrite(RGB,'newImage.png');
with:
sizeOfAcube the size of one cube on the QRcode.
numRows and numCols the number of Rows and Column of the original image.
One solution is to use patches, for instance as follows:
% set up example array
xl = 24; yl = xl;
[X Y] = find(hadamard(xl)==1);
% generate figure
figure, hold on
for ii=1:length(X)
patch(X(ii) + [0 0 1 1],Y(ii) + [0.1 0.9 0.9 0.1],[1 1 1],'Edgecolor',[1 1 1])
end
axis([0 xl+1 0 yl+1])
axis('square')
The patch command patch(x,y, color) accepts the vertices of the polygon element as x and y. In this example you can modify the term [0.1 0.9 0.9 0.1] to set the thickness of the bounding black line.
This generates
Edited
For the particular instance provided by the OP:
H=Hadamard(48); %# now to row-double the matrix
A=(1+H)/2;
B=(1-H)/2;
C=[A; B]; %# the code below randomly permutes elements within the rows of the matrix
[nRows,nCols] = size(C);
[junk,idx] = sort(rand(nRows,nCols),2); %# convert column indices into linear indices
idx = (idx-1)*nRows + ndgrid(1:nRows,1:nCols); %# rearrange whatever matrix
E = C;
E(:) = E(idx);
[X Y] = find(logical(E));
xl = length(X);
yl = length(Y);
figure, hold on
for ii=1:xl
rectangle('Position',[X(ii) Y(ii)+.2 1 0.8],'facecolor',[1 1 1],'edgecolor',[1 1 1])
end
axis([0 max(X)+1 0 max(Y)+1])
axis('square')
set(gca,'color',[0 0 0])
set(gca,'XTickLabel',[],'YTickLabel',[],'XTick',[],'YTick',[])
This example uses rectangle instead of patch to generate sharp corners.
The image:
I'd like to have a surface plot like the one below, but with a proper colorbar.
This is my code:
[X,Y,Z] = peaks(30);
[maxval dummy] = max(Z(:));
[minval dummy] = min(Z(:));
crange = 1.5;
% red, yellow, green
cmap = [1 0 0; 1 1 0; 0 1 0];
colormap(cmap);
colors = zeros(size(Z));
colors(Z <= -crange) = 1; % red (1)
colors(Z > -crange & Z < crange) = 2; % yellow (2)
colors(Z >= crange) = 3; % green (3)
surf(X,Y,Z, colors);
axis([-3 3 -3 3 -10 10]);
%cbh = colorbar('YGrid','on');
%caxis([minval-0.1 maxval+0.1]);
%set(cbh,'YTick',[minval -crange crange maxval]);
So far I had no luck in adding a colorbar where the colors (green,yellow,red) are aligned according to my custom range (green[8 ... 1.5],yellow[1.5 ... -1.5], red[-1.5 ... -6.4]). Instead, when I uncomment the last three lines,
a colorbar with linearly aligned colors shows up and the colors in my plot are aligned according to the colorbar and not to my custom range.
Now, what I'd like to have is that the colors in the colorbar match my custom ticks and that the plot looks like in the first picture.
The problem is that you specify the colors of each point in the surf plot by yourself, so they're not related to the z-value as is by default. The colorbar therefor is constructed only based on the color numbers, being 1 to 3. These therefor also show up as the default ticks of the colorbar (before you change them.
As you found out, you can set the ticks manually, and in the same way you can 'cheat' and use yticklabels:
figure
colormap(cmap);
surf(X,Y,Z, colors);
axis([-3 3 -3 3 -10 10]);
cbh = colorbar('YGrid','on');
set(cbh,'ytick',linspace(1,3,4));
set(cbh,'yticklabel',arrayfun(#num2str,[minval -crange crange maxval],'uni',false));
Or another way is to simply use caxis, but then the colors of the plot are linearly defined by the minmax values. So with this you can't set your non-linear ranges.
Illustration:
figure
colormap(cmap);
surf(X,Y,Z);
axis([-3 3 -3 3 -10 10]);
caxis([minval-0.1 maxval+0.1]);
cbh=colorbar
set(cbh,'YTick',[minval -crange crange maxval]);
So after all, I think my first method (using yticklabels) is the only way of doing what you want.
I know this is a crazy old post, but it came up as I was looking for answers. And here would be my answer (which has to assume unfortunately that the graduations on the color bar are equal size).
So given that you have made the color map, which is only 3 colors, the next part is algebra. The caxis follows a simple formula which is determined by the how many colors are in the color bar, and your min and max range.
index = fix((C-cmin)/(cmax-cmin)*m)+1;
So index will refer to the index in the color map (m = 3 since that is the length of your colormap), and what I would do is make C= 1.5, decide if you want it symmetrical (you are solving for cmax and cmin and is easier if both are x), and make index=2 (since you only have 3 colors, that should mean solving for 1.5 would give you the cmin/cmax to set which would make 1.5 the cutoff between yellow and red. You should be able to set the display range value somewhere, which will set the colormap appropriately (though probably not the labels).
Sorry to reply to such an old post, but maybe this will help others.
I came across this question when finding solutions for a problem that I had. Anyway this question helped me to get a solution for my problem after making little modifications on written codes.
At the same time I like to suggest some changes for the code so that user can get the required plot with necessary colors and the appropriate color bar also.
Here is the code,
[X,Y,Z] = peaks(30);
[maxval dummy] = max(Z(:));
[minval dummy] = min(Z(:));
crange=1.5;
% red, yellow, green
cmap = [1 0 0; 1 1 0; 0 1 0];
colormap(cmap);
colors = zeros(size(Z));
colors(Z <= -crange) = minval-0.1; % red (1)
colors(Z > -crange & Z < crange) = crange; % yellow (2)
colors(Z >= crange) = maxval+0.1; % green (3)
surf(X,Y,Z, colors);
axis([-3 3 -3 3 -10 10]);
cbh = colorbar('YGrid','on');
caxis([minval-0.1 maxval+0.1]);
set(cbh,'YTick',[minval -crange crange maxval]);
Plot