I'm trying to draw ROC curves using an existing set of values using the following command
plot(X1,Y1,'--rs',X2,Y2,'-*g');
Where X1,Y1,X2 and Y2 are matrices that have the same size
However, the lines produced by this command are straight ones.
How can I make them curved lines.
Thanks
Aziz
MATLAB by default uses straight line approximation to draw your graph in between control points. If you want, you can interpolate in between the points to produce a more realistic graph. Try using interp1 with the 'spline' option and see how that goes. As such, figure out the minimum and maximum values of both X1 and X2, then define a grid of points in between the minimum and maximum that have finer granularity. Once you do this, throw this into interp1 and plot your curve. Something like:
%// Find dynamic range of domain for both Xs
minX1 = min(X1);
maxX1 = max(X1);
minX2 = min(X2);
maxX2 = max(X2);
%// Generate grid of points for both Xs
x1Vals = linspace(minX1, maxX1, 100);
x2Vals = linspace(minX2, maxX2, 100);
%// Interpolate the curves
y1Vals = interp1(X1, Y1, x1Vals, 'spline');
y2Vals = interp1(X2, Y2, x2Vals, 'spline');
%// Plot the results
plot(x1Vals,y1Vals,'--rs',x2Vals,y2Vals,'-*g');
linspace generates a grid of points from one end to another, and I specified 100 of these points. I then use interp1 in the way we talked about where you specify control points (X1,Y1,X2,Y2), then specify the values I want to interpolate with. I use the output values after interpolation and draw the curve.
Related
I have this data with values on the edges of the matrix and other values at evenly spaced interval within the matrix. I want to predict the values of the zero positions from the original values and make a heat map of the new data. Through suggest, I use scatteredInterpolant, ndgrid and interpolant since the data is that interp2 (matlab functions) cannot be used to interpolate the zero elements. Now, this method doe not give me a smooth figure and I am want to know if someone can offer some help. I have attached the figure from my code, the data and the code to this post.Thank you.
[knownrows, knowncolumns, knownvalues] = find(DataGrid); %get location and value of all non-zero points
interpolant = scatteredInterpolant(knownrows, knowncolumns, knownvalues,'linear'); %create interpolant from known values
[queryrows, querycolumns] = ndgrid(1:1:size(DataGrid, 1), 1:1:size(DataGrid, 2)); %create grid of query points
interpolatedj = interpolant(queryrows, querycolumns);
HeatMap(interpolatedj)
https://www.mediafire.com/?pq40x1ljxk8h996
https://www.mediafire.com/?pq40x1ljxk8h996
To plot a smoothed matrix you can use pcolor and set the shading parameter to interp
pcolor(M); %where M is your 2D matrix
shading interp %set the shading to interp
Try
image(M) or imagesc(M) where M is a matrix. pcolor(M) also works. If your matrix is huge then you need to remove edges otherwise figure just looks like blank image.
Hope the title gave an adequate description of my problem. Basically, I am generating a contour plot in MATLAB using the contourf (x,y,z) function, where x and y are vectors of different lengths and z is a matrix of data with dimensions of x times y. The contourf plot is fine, however, I am looking to overlay this plot with the actual data points from the matrix z. I have tried using the scatter function, but I am getting an error message informing me that X and Y must be vectors of the same length - which they're not. Is there any other way to achieve this?
Thanks in advance for any help/suggestions!
I think meshgrid should help you.
z = peaks; %// example 49x49 z data
x = 1:20;
y = 1:49;
z = z(y,x); %// make dimensions not equal so length(x)~=length(y)
[c,h] = contourf(x,y,z);
clabel(c,h); colorbar;
[xx,yy]=meshgrid(x,y); %// this is what you need
hold on;
plot(xx,yy,'k.'); %// overlay points on contourf
Notice plot suffices instead of scatter. If you insist, scatter(xx(:),yy(:),10), for example, does the trick. Although my example isn't particularly interesting, this should hopefully get you started toward whatever you're going for aesthetically.
So I have data in the form [x y z intensity] that I plot on a scatter3 figure with xyz axes. The colour of the data is used to dictate the intensity value. Problem is, using a scatter plot means the data points show up as discrete points. What I need, is a smooth shape - so I guess I need some kind of interpolation between the points?
I've tried using trisurf, but the problem with this one is that it interpolates between points that it shouldn't. So where I should have 'gaps' in my surface, it joins up the edges instead so it fills in the gaps. See the attached pics for clarification.
Does anyone have any suggestions?
The code I use is as below (the commented out scatter3 is what does the scatter plot, the rest does the trisurf):
% Read in data
dataM = csvread('3dDispersion.csv');
% scatter3(dataM(:,1), dataM(:,2), dataM(:,3), 5, dataM(:,4),'filled');
% Plot
hold on;
x = dataM(:,1);
y = dataM(:,2);
freq = dataM(:,3);
tri = delaunay(x,y);
h = trisurf(tri, x, y, freq);
% Make it pretty
% view(-45,30);
view(3);
axis vis3d;
lighting phong;
shading interp;
Use the boundary function in Matlab. This will apply a mesh similar to shrinkwrap over your points. In order to reduce the "gap closers", you will want to increase the "shrink factor".
Try K = boundary(X,Y,Z,0.9)
Where X, Y & Z are the vectors of your data points
https://www.mathworks.com/help/matlab/ref/boundary.html
You can then use trimesh or related surface plotting functions depending on how you want to display it.
What I want to do is very simple, I just cannot seem to get MATLAB to do it. I would like to plot contours using my 2D data set.
My data set is large; 2 x 844240. I can do a scatter plot just fine,
scatter(Data(1,:), Data(2,:));
Reading through the forums I found Scatter plot with density in Matlab, where a hisogram was plotted. This would suffice, however, I would like to overlay the plots.
The issue is that they have different axis, my scatter data has an axis of [0 0.01 0 2500]; whereas the histogram is [0 100 0 100].
Is there a way to change the axis values of the histogram without modifying the image?
Thanks!
If I understand correctly, you are using hist3 to construct a histogram and then using imagesc to plot it. You can use the second output argument of hist3 to get the histogram bin centers, and then pass those on to imagesc, e.g.
nBins_x = 100;
nBins_y = 100;
[counts, bin_centers] = hist3(Data, [nBins_x nBins_y]);
x_bin_centers = bin_centers{1};
y_bin_centers = bin_centers{2};
imagesc(x_bin_centers, y_bin_centers, counts)
A couple other notes:
In your case, you will need to transpose your [2 x N] matrix when passing it to hist3, which expects an [N x 2] matrix.
imagesc puts the first axis (which I've been calling the "x" axis) on the vertical axis and the second on the horizontal axis. If you want to flip it, you can use:
imagesc(y_bin_centers, x_bin_centers, counts')
If you want to specify the histogram bins explicitly (e.g. to match your scatterplot) you can specify that in the arguments to hist3:
x_bin_centers = linspace(0, .01, 100);
y_bin_centers = linspace(0, 2500, 100);
counts = hist3(Data, {x_bin_centers, y_bin_centers};
And if you want a contour plot, you can use (note that contour takes the axes arguments in a different order than imagesc):
contour(x_bin_centers, y_bin_centers, counts');
If you are unhappy with the jaggedness of the contours, you may consider using a kernel density estimate instead of a histogram (check out ksdensity) (oops, looks like ksdensity is 1-D only. But there are File Exchange submissions for bivariate kernel density estimation).
I've got a series of XY point pairs in MATLAB. These pairs describe points around a shape in an image; they're not a function, meaning that two or more y points may exist for each x value.
I can plot these points individually using something like
plot(B(:,1),B(:,2),'b+');
I can also use plot to connect the points:
plot(B(:,1),B(:,2),'r');
What I'm trying to retrieve are my own point values I can use to connect the points so that I can use them for further analysis. I don't want a fully connected graph and I need something data-based, not just the graphic that plot() produces. I'd love to just have plot() generate these points (as it seems to do behind the scenes), but I've tried using the linseries returned by plot() and it either doesn't work as I understand it or just doesn't give me what I want.
I'd think this was an interpolation problem, but the points don't comprise a function; they describe a shape. Essentially, all I need are the points that plot() seems to calculate; straight lines connecting a series of points. A curve would be a bonus and would save me grief downstream.
How can I do this in MATLAB?
Thanks!
Edit: Yes, a picture would help :)
The blue points are the actual point values (x,y), plotted using the first plot() call above. The red outline is the result of calling plot() using the second approach above. I'm trying to get the point data of the red outline; in other words, the points connecting the blue points.
Adrien definitely has the right idea: define a parametric coordinate then perform linear interpolation on the x and y coordinates separately.
One thing I'd like to add is another way to define your parametric coordinate so you can create evenly-spaced interpolation points around the entire shape in one pass. The first thing you want to do, if you haven't already, is make sure the last coordinate point reconnects to the first by replicating the first point and adding it to the end:
B = [B; B(1,:)];
Next, by computing the total distance between subsequent points then taking the cumulative sum, you can get a parametric coordinate that makes small steps for points close together and larger steps for points far apart:
distance = sqrt(sum(diff(B,1,1).^2,2)); %# Distance between subsequent points
s = [0; cumsum(distance)]; %# Parametric coordinate
Now, you can interpolate a new set of points that are evenly spaced around the edge along the straight lines joining your points using the function INTERP1Q:
sNew = linspace(0,s(end),100).'; %'# 100 evenly spaced points from 0 to s(end)
xNew = interp1q(s,B(:,1),sNew); %# Interpolate new x values
yNew = interp1q(s,B(:,2),sNew); %# Interpolate new y values
These new sets of points won't necessarily include the original points, so if you want to be sure the original points also appear in the new set, you can do the following:
[sAll,sortIndex] = sort([s; sNew]); %# Sort all the parametric coordinates
xAll = [B(:,1); xNew]; %# Collect the x coordinates
xAll = xAll(sortIndex); %# Sort the x coordinates
yAll = [B(:,2); yNew]; %# Collect the y coordinate
yAll = yAll(sortIndex); %# Sort the y coordinates
EXAMPLE:
Here's an example to show how the above code performs (I use 11 pairs of x and y coordinates, one of which is repeated for the sake of a complete example):
B = [0.1371 0.1301; ... %# Sample data
0.0541 0.5687; ...
0.0541 0.5687; ... %# Repeated point
0.0588 0.5863; ...
0.3652 0.8670; ...
0.3906 0.8640; ...
0.4090 0.8640; ...
0.8283 0.7939; ...
0.7661 0.3874; ...
0.4804 0.1418; ...
0.4551 0.1418];
%# Run the above code...
plot(B(:,1),B(:,2),'b-*'); %# Plot the original points
hold on; %# Add to the plot
plot(xNew,yNew,'ro'); %# Plot xNew and yNew
I'd first define some parametric coordinate along the different segments (i.e. between the data points)
s = 1:size(B,1);
Then, just use interp1 to interpolate in s space. e.g. If you want to generate 10 values on the line between data point 5 and 6 :
s_interp = linspace(5,6,10); % parametric coordinate interpolation values
x_coord = interp1(s,B(:,1),s_interp,'linear');
y_coord = interp1(s,B(:,2),s_interp,'linear');
This should do the trick.
A.
Actually there is a MATLAB function "improfile", which might help you in your problem. Lets say these are the 4 coordinates which you want to find the locations between these coordinates.
xi=[15 30 20 10];
yi=[5 25 30 50];
figure;
plot(xi,yi,'r^-','MarkerSize',12)
grid on
Just generate a random image and run the function
n=50; % total number of points between initial coordinates
I=ones(max([xi(:);yi(:)]));
[cx,cy,c] = improfile(I,xi,yi,n);
hold on, plot(cx,cy,'bs-','MarkerSize',4)
Hope it helps