Quick count of weighted number of points inside grid cells

ParaView Support
1

Hi All,

I have a set of random points in 3D space and each one of them has a variable which describes its importance.

points
points840×900 298 KB

I use the FastUniformGrid and Transform filter to generate a bounding grid for the points as below:

grid
grid841×901 131 KB

NOTE the grid in the Y direction is just one cell

Grid in Y
Grid in Y839×897 109 KB

I would like to have on the grid cells the number of points they contain. Furthermore I would like to know the weighted number of points based on the importance variable.

Is there a quick way to do this with combination of filters?

Thanks
Fotos

2

Using the vtkCellLocator in the Programmable filter, you can count the number of points contained in the grid’s cells. The procedure is as follows:

  1. In the Pipeline Browser, select the grid dataset.
  2. Hold down the Crtl key and select the Point dataset.
  3. Call the Programmable Filter.
  4. Write the following in Script and Apply.
grid = self.GetInputDataObject(0, 0)
pnts = self.GetInputDataObject(0, 1)
output.CopyStructure(grid)

import paraview.vtk as vtk
import numpy as np

locator = vtk.vtkCellLocator()
locator.SetDataSet(grid)
locator.BuildLocator()

num_cells = output.GetNumberOfCells()
np_counts = np.zeros(num_cells, dtype=int)
num_pnts = pnts.GetNumberOfPoints()
for i in range(num_pnts):
    pt = pnts.GetPoint(i)
    cell_id = locator.FindCell(pt)
    np_counts[cell_id] += 1

output.CellData.append(np_counts, 'count')

count_pnts
count_pnts1318×825 243 KB

Upload the State file for the above.
count_pnts.pvsm (524.9 KB)

The weighted version can be achieved with a few modifications.

3

Thanks! perfect!