从非结构化点到非结构化点的2D插值

Question

I'm stuck at this things for last couple of days. I have two sets of unstructured data. First set of data has only 9583 points and second one has 60000 points. I need to interpolate a particular field of first set data(x1[len = 9583], y1[len = 9583], temperature1[len = 9583]) to second set of data (x2[len = 60000], y2[len = 60000], temperature2[len = 60000]). I have tried with python "SmoothBivariateSpline" and looked for other options. But I did not find any reasonable solutions. I'm open for either python or c++ . It would be highly appreciable if anybody can help me figuring out the solution. I attached the figure for your convenience Figure shows two sets of unstructured

Answer 1

There's a bunch on SO on unstructured interpolation, including:

1. Python/Scipy 2D Interpolation (Non-uniform Data)
2. What to do if I want 3D spline/smooth interpolation of random unstructured data?
3. 3D interpolation of 2 lists in python

And of course, there's always the SciPy Interpolation Documentation

Contrary to the title of your question, your data looks more like what Numpy calls "1D data" (that is XY data) rather than "2D data" (XYZ data). It also looks like your data is mostly sampling & noise, rather than fit to some mathematical function. So, I'd suggest starting at the top of the list from the SciPy Docs , skipping over the univariate and multivariate stuff and trying something in the 1-D Splines section, perhaps Scipy.interpolate.BSpline .