I have a prescribed velocity field on a uniform, collocated 2D grid. I need to compute the divergence of the velocity field at some arbitrary points that does not coincide with the grid points or midpoint location.
On posibility could be to compute the divergence on the grid points and then interpolate using bicubic interpolation at the unknown points. I am not sure this is a good idea because it could create some checkerboard effect.
I wonder if there is a better way to do that.
I am using python/scipy.
Related
I would like to find residuals of my solutions. However, each solution uses a different grid, thus I would like to interpolate the solutions in 2D to a mesh grid for each solution such that I can easily find the residuals.
I used griddata which seems to work, but it doesn't work very well in regions of discontinuities and near edges. I would like to interpolate data of flow over a sphere for example. The area of the sphere should be confined to zero since that is outside of the grid boundaries; however, after using griddata, there's some non zero values in the sphere region. I also get a warning saying there were duplicate x and y points.
Does anyone know how to preserve the integrity of the sphere after using griddata? I've double checked the original x,y data, and the area of the sphere all have values of 0.
Thank you. I hope that made sense.
Best,
Yuki
So I have a 3 dimensional matrix of points that (presumably) define a surface. For my purposes, X and Y can be random values but when plotted along with their Z coordinates, they will define some undulating surface. I'd like to measure the local curvatures of said surface, and in order to do that, I need to be able to find the gradient of said surface, at which point calculating the curvature is trivial.
I have not yet found an implementation of how to measure this curvature that doesn't make use of Matlab's gradient function. The problem with Matlab's gradient function is that it assumes that the points are in some sort of order, similar to diff(X). This would suffice if my points were spaced along a grid, which is not necessarily the case.
One possible solution to measuring the gradient is to give in and assign each point to a discrete coordinate in a grid in the XY plane, thus overcoming this issue. However, this solution seems somewhat inelegant and was curious to see if anyone had suggestions. Thanks!
You can use griddata to interpolate from your scattered data points to grid spaced points and then calculate the gradient.
I have N 3D observations taken from an optical motion capture system in XYZ form.
The motion that was captured was just a simple circle arc, derived from a rigid body with fixed axis of rotation.
I used the princomp function in matlab to get all marker points on the same plane i.e. the plane on which the motion has been done.
(See a pic representing 3D data on the plane that was found, below)
What i want to do after the previous step is to look the fitted data on the plane that was found and get the curve of the captured motion in 2D.
In the princomp how to, it is said that
The first two coordinates of the principal component scores give the
projection of each point onto the plane, in the coordinate system of
the plane.
(from "Fitting an Orthogonal Regression Using Principal Components Analysis" article on mathworks help site)
So i thought that if i just plot those pc scores -plot(score(:,1),score(:,2))- i'll get the motion curve. Instead what i got is this.
(See a pic representing curve data in 2D derived from pc scores, below)
The 2d curve seems stretched and nonlinear (different y values for same x values) when it shouldn't be. The curve that i am looking for, should be interpolated by just using simple polynomial (polyfit) or circle fit in matlab.
Is this happening because the plane that was found looks like rhombus relative to the original coordinate system and the pc axes are rotated with respect to the basis of plane in such way that produce this stretch?
Then i thought that, this is happening because of the different coordinate systems of optical system and Matlab. Optical system's (ie cameras) co.sys. is XZY oriented and Matlab's default (i think) co.sys is XYZ oriented. I transformed my data to correspond to Matlab's co.sys through a rotation matrix, run again princomp but i got the same stretch in the 2D curve (the new curve just had different orientation now).
Somewhere else i read that
Principal Components Analysis chooses the first PCA axis as that line
that goes through the centroid, but also minimizes the square of the
distance of each point to that line. Thus, in some sense, the line is
as close to all of the data as possible. Equivalently, the line goes
through the maximum variation in the data. The second PCA axis also
must go through the centroid, and also goes through the maximum
variation in the data, but with a certain constraint: It must be
completely uncorrelated (i.e. at right angles, or "orthogonal") to PCA
axis 1.
I know that i am missing something but i have a problem understanding why i get a stretched curve. What i have to do so i can get the curve right?
Thanks in advance.
EDIT: Here is a sample data file (3 columns XYZ coords for 2 markers)
w w w.sendspace.com/file/2hiezc
I want to know how griddata uses linear interpolation.
Does it just consider like four neighbor points of the point where we want to find the value or uses the whole convex hull?
The docs for griddata will guide you in this. Essentially, it does a delaunay triangulation of your point set, then for any new point, determine which simplex the point falls in. Interpolation within a simplex is linear. Thus in 2 dimensions, three points define a triangle, and 3 points determine a locally planar model for z(x,y).
I have a formula that depends on theta and phi (spherical coordinates 0<=theta<=2*pi and 0<=phi<=pi). By inserting each engle, I obtained a quantity. Now I have a set of data for different angles and I need to plot the surface. My data is a 180*360 matrix, so I am not sure if I can use SURF or MESH or PLOT3. The figure should be a surface that include all data and the axes should be in terms of the quantity, not the quantity versus the angles. How can I plot such a surface?
I see no reason why you cannot use mesh or surf to plot such data. Another option I tend to use is that of density plots. You basically display the dependent variable (quantity) as an image and include the independent variables (angles) along the axis, much like you would with the aforementioned 3D plotting functions. This can be done with imagesc.
Typically you would want your axes to be the dependent variables. Could you elaborate more on this point?
If I understand you correctly you have calculated a function f(theta,phi) and now you want to plot the surface containing all the points with the polar coordinated (r,theta,phi) where r=f(theta,phi).
If this is what you want to do, the 2D version of such a plot is included in MATLAB under the name polar. Unfortunately, as you pointed out, polar3 on MatlabCentral is not the generalization you are looking for.
I have been able to plot a sphere with the following code, using constant r=1. You can give it a try with your function:
phi1=0:1/(3*pi):pi; %# this would be your 180 points
theta1=-pi:1/(3*pi):pi; % your 360 points
r=ones(numel(theta1),numel(phi1));
[phi,theta]=meshgrid(phi1,theta1);
x=r.*sin(theta).*cos(phi);
y=r.*sin(theta).*sin(phi);
z=r.*cos(theta);
tri=delaunay(x(:),y(:),z(:));
trisurf(tri,x,y,z);
From my tests it seems that delaunay also includes a lot of triangles which go through the volume of my sphere, so it seems this is not optimal. So maybe you can have a look at fill3 and construct the triangles it draws itself: as a first approximation, you could have the points [x(n,m) x(n+1,m) x(n,m+1)] combined into one triangle, and [x(n+1,m) x(n+1,m+1) x(n+1,m+1)] into another...?