extract values at nearest point with multidimensional latitude and longitude field

[jinhuanz]

How do I extract values at a specific latitude and longitude point when latitude/longitude values are 2-D array?

I have a dataset with the following coordinates:

Dimensions: (x: 6165, y: 5801)
Coordinates:

• y (y) float64 0.0 3e+03 6e+03 9e+03 1.2e+04 1.5e+04 ...
• x (x) float64 0.0 3e+03 6e+03 9e+03 1.2e+04 1.5e+04 ...
  latitude (y, x) float64 10.16 10.17 10.17 10.18 10.19 10.19 ...
  longitude (y, x) float64 78.99 79.01 79.03 79.05 79.08 79.1 ...

I want to extract value for at a specific latitude and longitude point and the exact latitude/longitude value is not in the coordinates. Then I started with calculate the distance between two point on Earth, and then get the (x_idx, y_idx) by minimize the distance. However, CPU and time costing is a result because of the large size of data.

I'm wondering is there any other ways could make it less costing?

[shoyer]

The right solution is to use a KDTree. We've discussed possible syntax for this in #475 but nobody has implemented it yet. Russ Rew wrote a nice overview article on using KDTree in http://www.unidata.ucar.edu/blogs/developer/entry/accessing_netcdf_data_by_coordinates

[stale]

In order to maintain a list of currently relevant issues, we mark issues as stale after a period of inactivity

If this issue remains relevant, please comment here or remove the stale label; otherwise it will be marked as closed automatically