improver.regrid.nearest module
Nearest neighbour interpolation functions
- lakes_islands(lake_island_indexes, indexes, surface_type_mask, in_latlons, out_latlons, in_classified, out_classified, vicinity)[source]
Updating source points and weighting for 4-unmatching-source-point cases - water surrounded by land or land surrounded by water. This function searches nearest 8 points to check if any matching point exists. Note that a similar function can be found in bilinear.py for bilinear regridding rather than nearest neighbour regridding.
- Parameters:
lake_island_indexes (
ndarray) – Indexes of points which are lakes/islands surrounded by mismatched surface type. These points will be processed by this function.in_latlons (
ndarray) – Source points’s latitude-longitudes.out_latlons (
ndarray) – Target points’s latitude-longitudes.surface_type_mask (
ndarray) – Boolean true if source point type matches target point type.indexes (
ndarray) – Source grid point indexes for each target grid point.in_classified (
ndarray) – Land/sea type for source grid points (land -> True).out_classified (
ndarray) – Land/sea type for target grid points (land -> True).vicinity (
float) – Radius of vicinity to search for a matching surface type, in metres.
- Return type:
- Returns:
Updated indexes - source grid point number for all target grid points.
Updated surface_type_mask - matching info between source/target point types.
- nearest_regrid(distances, indexes, in_values)[source]
Main regridding function for the nearest neighbour option.
- Parameters:
- Return type:
- Returns:
Regridded output values with spatial dimensions flattened.
- nearest_with_mask_regrid(distances, indexes, surface_type_mask, in_latlons, out_latlons, in_classified, out_classified, vicinity)[source]
Main regridding function for the nearest distance option. some input just for handling island-like points.
- Parameters:
distances (
ndarray) – Distnace array from each target grid point to its source grid points.indexes (
ndarray) – Source grid point indexes for each target grid point.surface_type_mask (
ndarray) – Boolean true if source point type matches target point type.in_latlons (
ndarray) – Source points’s latitude-longitudes.out_latlons (
ndarray) – Target points’s latitude-longitudes.in_classified (
ndarray) – Land/sea type for source grid points (land -> True).out_classified (
ndarray) – Land/sea type for target grid points (land -> True).vicinity (
float) – Radius of specified searching domain, in meter.
- Return type:
- Returns:
Updated distances - array from each target grid point to its source grid points.
Updated indexes - source grid point number for all target grid points.
- update_nearest_points(points_with_mismatches, in_latlons, out_latlons, indexes, distances, surface_type_mask, in_classified, out_classified)[source]
Update nearest source points and distances/surface_type to take into account surface type of nearby points.
- Parameters:
points_with_mismatches (
ndarray) – Selected target points which will use Inverse Distance Weighting (idw) approach. These points will be processed by this function.in_latlons (
ndarray) – Source points’s latitude-longitudes.out_latlons (
ndarray) – Target points’s latitude-longitudes.indexes (
ndarray) – Source grid point indexes for each target grid point.distances (
ndarray) – Distance from each target grid point to its source grid points.surface_type_mask (
ndarray) – Boolean true if source point type matches target point type.in_classified (
ndarray) – Land/sea type for source grid points (land -> True).out_classified (
ndarray) – Land/sea type for target grid points (land -> True).
- Return type:
- Returns:
Updated indexes - source grid point number for all target grid points.
Updated distances - array from each target grid point to its source grid points.
Updated surface_type_mask - matching info between source/target point types.