1

I have xyz coordinates with some associated data (classification and angle). I would like to use numpy.meshgrid to use a function to summarize the point data within each grid cell (e.g. mean z, std, etc.). The reason I would like to use numpy.meshgrid is for scalability (geopandas is not feasible for my project, since we are using a point cloud that contains millions of points).

So, I simply want to be able to go through each grid cell, filter for points within that grid cell (see data below e.g. lasx, lasy, and lasz) and compute mean, standard deviation, count, etc. (e.g. of the z-values below).

Here is my mesh grid:

binx=np.arange(min(lasx),max(lasx),30)
biny=np.arange(min(lasy),max(lasy),30)

xx, yy = np.meshgrid(binx,biny)

Here is some data for testing:

lasx = np.array([320999.12, 320997.33, 320995.03, 320996.22, 320995.61, 320996.65,
       320991.71, 320992.75, 320991.95, 320992.89, 320991.23, 320994.58,
       320992.58, 320985.69, 320987.6 , 320998.81, 320984.84, 320990.01,
       320991.56, 320983.47, 320985.07, 320988.36, 320984.21, 320997.23,
       320982.59, 320994.63, 320983.21, 320985.7 , 320995.28, 320984.52,
       320995.12, 320981.91, 320994.09, 320985.04, 320994.28, 320984.22,
       320990.29, 320985.68, 320997.12, 320984.28, 320994.98, 320984.6 ,
       320996.17, 320984.68, 320998.33, 320984.47, 320998.41, 320986.34,
       320996.67, 320992.11, 320987.13, 320998.71, 320987.07, 320995.59,
       320988.23, 320994.73, 320994.73, 320989.05, 320995.25, 320990.47,
       320994.11, 320996.89, 320992.02, 320992.08, 320993.  , 320993.72,
       320994.44, 320995.68, 320996.37, 320999.77, 320999.09, 320999.94,
       320717.88, 320732.82, 320753.15, 320719.04, 320726.18, 320770.27,
       320825.5 , 320779.35, 320751.4 , 320717.6 , 320731.44, 320764.28,
       320798.3 , 320852.99, 320860.2 , 320814.51, 320774.91, 320738.15,
       320718.73, 320749.17, 320777.43, 320814.38, 320857.39, 320911.5 ,
       320918.48, 320876.19, 320836.37, 320786.07])

lasy = np.array([4096922.47, 4096925.41, 4096926.95, 4096928.31, 4096931.07,
       4096932.53, 4096932.82, 4096933.65, 4096934.99, 4096936.38,
       4096938.45, 4096939.61, 4096939.75, 4096941.53, 4096942.79,
       4096942.22, 4096944.24, 4096945.29, 4096945.26, 4096947.01,
       4096948.2 , 4096948.05, 4096949.32, 4096947.87, 4096950.99,
       4096950.31, 4096952.26, 4096952.13, 4096952.25, 4096953.25,
       4096953.56, 4096954.96, 4096954.27, 4096955.59, 4096956.57,
       4096957.53, 4096958.38, 4096958.17, 4096959.27, 4096960.12,
       4096960.09, 4096961.46, 4096961.96, 4096962.74, 4096962.57,
       4096964.1 , 4096964.47, 4096965.3 , 4096964.93, 4096966.23,
       4096967.9 , 4096968.43, 4096969.24, 4096969.63, 4096970.49,
       4096970.53, 4096971.41, 4096973.11, 4096973.5 , 4096974.32,
       4096975.09, 4096975.27, 4096976.93, 4096978.23, 4096979.54,
       4096980.75, 4096982.02, 4096983.37, 4096985.85, 4096987.05,
       4096989.59, 4096993.2 , 4096999.77, 4096999.86, 4096999.47,
       4096998.85, 4096998.77, 4096999.45, 4096999.48, 4096998.73,
       4096999.43, 4096997.65, 4096997.91, 4096999.43, 4096998.9 ,
       4096999.71, 4096999.17, 4096998.4 , 4096998.27, 4096997.16,
       4096996.74, 4096997.86, 4096997.69, 4096998.24, 4096999.32,
       4096999.82, 4096999.17, 4096998.39, 4096997.68, 4096996.87])

lasz = np.array([ 0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  8.83,  8.86,
        9.29,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,
        0.  ,  0.  ,  0.  ,  0.  ,  5.36, 16.22,  0.21,  0.  ,  0.  ,
        0.  ,  9.89,  4.94,  9.14,  0.  ,  0.  ,  8.53,  4.82,  0.05,
        0.  ,  8.24,  0.  ,  0.  , 10.74,  0.  ,  0.  ,  0.  ,  9.61,
        0.  ,  0.  ,  0.  , 14.81,  0.  ,  0.  ,  0.  ,  0.  ,  5.08,
        0.  , 13.13,  0.  ,  0.  ,  7.39,  0.  ,  5.48,  0.  ,  0.  ,
        0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,  0.  ,
        3.  ,  0.  ,  3.51,  3.28,  0.  ,  0.  ,  0.  ,  0.  , 18.37,
        0.  ,  0.  , 16.26,  0.08,  0.  ,  0.  ,  0.25,  8.08,  0.  ,
        0.  ,  9.44,  0.  ,  0.26,  6.45,  0.  ,  0.  ,  0.  ,  0.  ,
        0.  ])
Clouseau
  • 27
  • 5
  • I'd look into SciPy's KDtree – norok2 Nov 09 '21 at 17:18
  • @norok2 Thanks for the response, but SciPy's KDtree is for nearest neighbor search. Does not provide any grid data functionality that allows for filtering within each grid cell. – Clouseau Nov 09 '21 at 17:52
  • I am not sure what you are after, but KDTree can be queried for points inside a certain radius, with abs distance that is exactly finding the points inside a "grid cell". Or am I missing something? – norok2 Nov 10 '21 at 10:31
  • Could you provide an example? I can query within a radius but I am not able to search for only absolute distance. – Clouseau Nov 11 '21 at 23:38

0 Answers0