def test_iter_spatial_decomposition(self):
self.add_barrier = False
if vm.size not in [1, 4]:
raise SkipTest('vm.size not in [1, 4]')
grid = create_gridxy_global(resolution=10.,
wrapped=False,
crs=Spherical())
splits = (2, 3)
actual = []
for sub, slc in iter_spatial_decomposition(grid,
splits,
optimized_bbox_subset=True):
root = vm.get_live_ranks_from_object(sub)[0]
with vm.scoped_by_emptyable('test extent', sub):
if vm.is_null:
extent_global = None
else:
extent_global = sub.extent_global
extent_global = vm.bcast(extent_global, root=root)
actual.append(extent_global)
desired = [(0.0, -90.0, 120.0, 0.0), (120.0, -90.0, 240.0, 0.0),
(240.0, -90.0, 360.0, 0.0), (0.0, 0.0, 120.0, 90.0),
(120.0, 0.0, 240.0, 90.0), (240.0, 0.0, 360.0, 90.0)]
self.assertEqual(actual, desired)
def test_iter_spatial_decomposition(self):
self.add_barrier = False
if vm.size not in [1, 4]:
raise SkipTest('vm.size not in [1, 4]')
grid = create_gridxy_global(resolution=10., wrapped=False, crs=Spherical())
splits = (2, 3)
actual = []
for sub, slc in iter_spatial_decomposition(grid, splits, optimized_bbox_subset=True):
root = vm.get_live_ranks_from_object(sub)[0]
with vm.scoped_by_emptyable('test extent', sub):
if vm.is_null:
extent_global = None
else:
extent_global = sub.extent_global
extent_global = vm.bcast(extent_global, root=root)
actual.append(extent_global)
desired = [(0.0, -90.0, 120.0, 0.0), (120.0, -90.0, 240.0, 0.0), (240.0, -90.0, 360.0, 0.0),
(0.0, 0.0, 120.0, 90.0), (120.0, 0.0, 240.0, 90.0), (240.0, 0.0, 360.0, 90.0)]
self.assertEqual(actual, desired)
def write_chunks(self):
"""
Write grid subsets to netCDF files using the provided filename templates. This will also generate ESMF
regridding weights for each subset if requested.
"""
src_filenames = []
dst_filenames = []
wgt_filenames = []
dst_slices = []
src_slices = []
index_path = self.create_full_path_from_template('index_file')
# nzeros = len(str(reduce(lambda x, y: x * y, self.nchunks_dst)))
ctr = 1
ocgis_lh(logger=_LOCAL_LOGGER,
msg='starting self.iter_src_grid_subsets',
level=logging.DEBUG)
for sub_src, src_slc, sub_dst, dst_slc in self.iter_src_grid_subsets(
yield_dst=True):
ocgis_lh(
logger=_LOCAL_LOGGER,
msg='finished iteration {} for self.iter_src_grid_subsets'.
format(ctr),
level=logging.DEBUG)
src_path = self.create_full_path_from_template('src_template',
index=ctr)
dst_path = self.create_full_path_from_template('dst_template',
index=ctr)
wgt_path = self.create_full_path_from_template('wgt_template',
index=ctr)
src_filenames.append(os.path.split(src_path)[1])
dst_filenames.append(os.path.split(dst_path)[1])
wgt_filenames.append(wgt_path)
dst_slices.append(dst_slc)
src_slices.append(src_slc)
# Only write destinations if an iterator is not provided.
if self.iter_dst is None:
zip_args = [[sub_src, sub_dst], [src_path, dst_path]]
else:
zip_args = [[sub_src], [src_path]]
cc = 1
for target, path in zip(*zip_args):
with vm.scoped_by_emptyable('field.write' + str(cc), target):
if not vm.is_null:
ocgis_lh(logger=_LOCAL_LOGGER,
msg='write_chunks:writing: {}'.format(path),
level=logging.DEBUG)
field = Field(grid=target)
field.write(path)
ocgis_lh(
logger=_LOCAL_LOGGER,
msg='write_chunks:finished writing: {}'.format(
path),
level=logging.DEBUG)
cc += 1
# Increment the counter outside of the loop to avoid counting empty subsets.
ctr += 1
# Generate an ESMF weights file if requested and at least one rank has data on it.
if self.genweights and len(
vm.get_live_ranks_from_object(sub_src)) > 0:
vm.barrier()
ocgis_lh(logger=_LOCAL_LOGGER,
msg='write_chunks:writing esmf weights: {}'.format(
wgt_path),
level=logging.DEBUG)
self.write_esmf_weights(src_path,
dst_path,
wgt_path,
src_grid=sub_src,
dst_grid=sub_dst)
vm.barrier()
# Global shapes require a VM global scope to collect.
src_global_shape = global_grid_shape(self.src_grid)
dst_global_shape = global_grid_shape(self.dst_grid)
# Gather and collapse source slices as some may be empty and we write on rank 0.
gathered_src_grid_slice = vm.gather(src_slices)
if vm.rank == 0:
len_src_slices = len(src_slices)
new_src_grid_slice = [None] * len_src_slices
for idx in range(len_src_slices):
for rank_src_grid_slice in gathered_src_grid_slice:
if rank_src_grid_slice[idx] is not None:
new_src_grid_slice[idx] = rank_src_grid_slice[idx]
break
src_slices = new_src_grid_slice
with vm.scoped('index write', [0]):
if not vm.is_null:
dim = Dimension('nfiles', len(src_filenames))
vname = [
'source_filename', 'destination_filename',
'weights_filename'
]
values = [src_filenames, dst_filenames, wgt_filenames]
grid_chunker_destination = GridChunkerConstants.IndexFile.NAME_DESTINATION_VARIABLE
attrs = [{
'esmf_role': 'grid_chunker_source'
}, {
'esmf_role': grid_chunker_destination
}, {
'esmf_role': 'grid_chunker_weights'
}]
vc = VariableCollection()
grid_chunker_index = GridChunkerConstants.IndexFile.NAME_INDEX_VARIABLE
vidx = Variable(name=grid_chunker_index)
vidx.attrs['esmf_role'] = grid_chunker_index
vidx.attrs['grid_chunker_source'] = 'source_filename'
vidx.attrs[GridChunkerConstants.IndexFile.
NAME_DESTINATION_VARIABLE] = 'destination_filename'
vidx.attrs['grid_chunker_weights'] = 'weights_filename'
vidx.attrs[GridChunkerConstants.IndexFile.
NAME_SRC_GRID_SHAPE] = src_global_shape
vidx.attrs[GridChunkerConstants.IndexFile.
NAME_DST_GRID_SHAPE] = dst_global_shape
vc.add_variable(vidx)
for idx in range(len(vname)):
v = Variable(name=vname[idx],
dimensions=dim,
dtype=str,
value=values[idx],
attrs=attrs[idx])
vc.add_variable(v)
bounds_dimension = Dimension(name='bounds', size=2)
# TODO: This needs to work with four dimensions.
# Source -----------------------------------------------------------------------------------------------
self.src_grid._gc_create_index_bounds_(RegriddingRole.SOURCE,
vidx, vc, src_slices,
dim, bounds_dimension)
# Destination ------------------------------------------------------------------------------------------
self.dst_grid._gc_create_index_bounds_(
RegriddingRole.DESTINATION, vidx, vc, dst_slices, dim,
bounds_dimension)
vc.write(index_path)
vm.barrier()
def iter_src_grid_subsets(self, yield_dst=False, yield_idx=None):
"""
Yield source grid subset using the extent of its associated destination grid subset.
:param bool yield_dst: If ``True``, yield the destination subset as well as the source grid subset.
:param int yield_idx: If a zero-based integer, only yield for this chunk index and skip everything else.
:rtype: tuple(:class:`ocgis.spatial.grid.AbstractGrid`, `slice-like`)
"""
if yield_dst:
yield_slice = True
else:
yield_slice = False
buffer_value = self.buffer_value
dst_grid_wrapped_state = self.dst_grid.wrapped_state
dst_grid_crs = self.dst_grid.crs
# Use a destination grid iterator if provided.
if self.iter_dst is not None:
iter_dst = self.iter_dst(self,
yield_slice=yield_slice,
yield_idx=yield_idx)
else:
iter_dst = self.iter_dst_grid_subsets(yield_slice=yield_slice,
yield_idx=yield_idx)
# Loop over each destination grid subset.
ocgis_lh(logger=_LOCAL_LOGGER,
msg='starting "for yld in iter_dst"',
level=logging.DEBUG)
for iter_dst_ctr, yld in enumerate(iter_dst, start=1):
ocgis_lh(msg=["iter_dst_ctr", iter_dst_ctr], level=logging.DEBUG)
if yield_slice:
dst_grid_subset, dst_slice = yld
else:
dst_grid_subset = yld
# All masked destinations are very problematic for ESMF
with vm.scoped_by_emptyable('global mask', dst_grid_subset):
if not vm.is_null:
if dst_grid_subset.has_mask_global:
if dst_grid_subset.has_mask and dst_grid_subset.has_masked_values:
all_masked = dst_grid_subset.get_mask().all()
else:
all_masked = False
all_masked_gather = vm.gather(all_masked)
if vm.rank == 0:
if all(all_masked_gather):
exc = ValueError(
"Destination subset all masked")
try:
raise exc
finally:
vm.abort(exc=exc)
dst_box = None
with vm.scoped_by_emptyable('extent_global', dst_grid_subset):
if not vm.is_null:
# Use the extent of the polygon for determining the bounding box. This ensures conservative
# regridding will be fully mapped.
if isinstance(dst_grid_subset,
AbstractGeometryCoordinates):
target_grid = dst_grid_subset.parent.grid
else:
target_grid = dst_grid_subset
# Try to reduce the coordinates in the case of unstructured grid data. Ensure the data also has a
# coordinate index. SCRIP grid files, for example, do not have a coordinate index like UGRID.
if hasattr(
target_grid, 'reduce_global'
) and Topology.POLYGON in target_grid.abstractions_available and target_grid.cindex is not None:
ocgis_lh(
logger=_LOCAL_LOGGER,
msg='starting reduce_global for dst_grid_subset',
level=logging.DEBUG)
target_grid = target_grid.reduce_global()
ocgis_lh(
logger=_LOCAL_LOGGER,
msg='finished reduce_global for dst_grid_subset',
level=logging.DEBUG)
extent_global = target_grid.parent.attrs.get(
'extent_global')
if extent_global is None:
with grid_abstraction_scope(target_grid,
Topology.POLYGON):
extent_global = target_grid.extent_global
# HACK: Bad corner coordinates can lead to bad extents. In this case, the lower bound on the
# x-coordinate is unreasonable and breaks wrapping code. Set to 0.0 which is a reasonable
# lower x-coordate for unwrapped datasets.
if (isinstance(target_grid.crs, Spherical)) and \
dst_grid_wrapped_state == WrappedState.UNWRAPPED and \
extent_global[0] < 0.0:
e = list(extent_global)
e[0] = 0.0
extent_global = tuple(e)
if self.check_contains:
dst_box = box(*target_grid.extent_global)
sub_box = box(*extent_global)
if buffer_value is not None:
# Use the envelope! A buffer returns "fancy" borders. We just want to expand the bounding box.
sub_box = sub_box.buffer(buffer_value).envelope
ocgis_lh(msg=str(sub_box.bounds),
level=logging.DEBUG,
logger=_LOCAL_LOGGER)
else:
sub_box, dst_box = [None, None]
live_ranks = vm.get_live_ranks_from_object(dst_grid_subset)
sub_box = vm.bcast(sub_box, root=live_ranks[0])
if self.check_contains:
dst_box = vm.bcast(dst_box, root=live_ranks[0])
sub_box = GeometryVariable.from_shapely(
sub_box,
is_bbox=True,
wrapped_state=dst_grid_wrapped_state,
crs=dst_grid_crs)
# Prepare geometry to match coordinate system and wrapping of the subset target
sub_box = sub_box.prepare(archetype=self.src_grid)
ocgis_lh(logger=_LOCAL_LOGGER,
msg='prepared geometry',
level=logging.DEBUG)
ocgis_lh(logger=_LOCAL_LOGGER,
msg='starting "self.src_grid.get_intersects"',
level=logging.DEBUG)
src_grid_subset, src_grid_slice = self.src_grid.get_intersects(
sub_box,
keep_touches=False,
cascade=False,
optimized_bbox_subset=self.optimized_bbox_subset,
return_slice=True)
ocgis_lh(logger=_LOCAL_LOGGER,
msg='finished "self.src_grid.get_intersects"',
level=logging.DEBUG)
# Reload the data using a new source index distribution.
if hasattr(src_grid_subset,
'reduce_global') and src_grid_subset.cindex is not None:
# Only redistribute if we have one live rank.
if self.redistribute and len(
vm.get_live_ranks_from_object(src_grid_subset)) > 0:
ocgis_lh(logger=_LOCAL_LOGGER,
msg='starting redistribute',
level=logging.DEBUG)
topology = src_grid_subset.abstractions_available[
Topology.POLYGON]
cindex = topology.cindex
redist_dimname = self.src_grid.abstractions_available[
Topology.POLYGON].element_dim.name
if src_grid_subset.is_empty:
redist_dim = None
else:
redist_dim = topology.element_dim
redistribute_by_src_idx(cindex, redist_dimname, redist_dim)
ocgis_lh(logger=_LOCAL_LOGGER,
msg='finished redistribute',
level=logging.DEBUG)
with vm.scoped_by_emptyable('src_grid_subset', src_grid_subset):
if not vm.is_null:
if not self.allow_masked:
gmask = src_grid_subset.get_mask()
if gmask is not None and gmask.any():
raise ValueError(
'Masked values in source grid subset.')
if self.check_contains:
src_box = box(*src_grid_subset.extent_global)
if not does_contain(src_box, dst_box):
raise ValueError('Contains check failed.')
# Try to reduce the coordinates in the case of unstructured grid data.
if hasattr(src_grid_subset, 'reduce_global'
) and src_grid_subset.cindex is not None:
ocgis_lh(logger=_LOCAL_LOGGER,
msg='starting reduce_global',
level=logging.DEBUG)
src_grid_subset = src_grid_subset.reduce_global()
ocgis_lh(logger=_LOCAL_LOGGER,
msg='finished reduce_global',
level=logging.DEBUG)
else:
pass
# src_grid_subset = VariableCollection(is_empty=True)
if src_grid_subset.is_empty:
src_grid_slice = None
else:
src_grid_slice = {
src_grid_subset.dimensions[ii].name: src_grid_slice[ii]
for ii in range(src_grid_subset.ndim)
}
if yield_dst:
yld = (src_grid_subset, src_grid_slice, dst_grid_subset,
dst_slice)
else:
yld = src_grid_subset, src_grid_slice
yield yld
def global_grid_shape(grid):
with vm.scoped_by_emptyable('global grid shape', grid):
if not vm.is_null:
return grid.shape_global
def iter_src_grid_subsets(self, yield_dst=False, yield_idx=None):
"""
Yield source grid subset using the extent of its associated destination grid subset.
:param bool yield_dst: If ``True``, yield the destination subset as well as the source grid subset.
:param int yield_idx: If a zero-based integer, only yield for this chunk index and skip everything else.
:rtype: tuple(:class:`ocgis.spatial.grid.AbstractGrid`, `slice-like`)
"""
if yield_dst:
yield_slice = True
else:
yield_slice = False
buffer_value = self.buffer_value
dst_grid_wrapped_state = self.dst_grid.wrapped_state
dst_grid_crs = self.dst_grid.crs
# Use a destination grid iterator if provided.
if self.iter_dst is not None:
iter_dst = self.iter_dst(self,
yield_slice=yield_slice,
yield_idx=yield_idx)
else:
iter_dst = self.iter_dst_grid_subsets(yield_slice=yield_slice,
yield_idx=yield_idx)
# Loop over each destination grid subset.
ocgis_lh(logger='grid_chunker',
msg='starting "for yld in iter_dst"',
level=logging.DEBUG)
for yld in iter_dst:
if yield_slice:
dst_grid_subset, dst_slice = yld
else:
dst_grid_subset = yld
dst_box = None
with vm.scoped_by_emptyable('extent_global', dst_grid_subset):
if not vm.is_null:
# Use the extent of the polygon for determining the bounding box. This ensures conservative
# regridding will be fully mapped.
if isinstance(dst_grid_subset,
AbstractGeometryCoordinates):
target_grid = dst_grid_subset.parent.grid
else:
target_grid = dst_grid_subset
extent_global = target_grid.parent.attrs.get(
'extent_global')
if extent_global is None:
with grid_abstraction_scope(target_grid,
Topology.POLYGON):
extent_global = target_grid.extent_global
if self.check_contains:
dst_box = box(*target_grid.extent_global)
sub_box = box(*extent_global)
if buffer_value is not None:
# Use the envelope! A buffer returns "fancy" borders. We just want to expand the bounding box.
sub_box = sub_box.buffer(buffer_value).envelope
ocgis_lh(msg=str(sub_box.bounds),
level=logging.DEBUG,
logger='grid_chunker')
else:
sub_box, dst_box = [None, None]
live_ranks = vm.get_live_ranks_from_object(dst_grid_subset)
sub_box = vm.bcast(sub_box, root=live_ranks[0])
if self.check_contains:
dst_box = vm.bcast(dst_box, root=live_ranks[0])
sub_box = GeometryVariable.from_shapely(
sub_box,
is_bbox=True,
wrapped_state=dst_grid_wrapped_state,
crs=dst_grid_crs)
ocgis_lh(logger='grid_chunker',
msg='starting "self.src_grid.get_intersects"',
level=logging.DEBUG)
src_grid_subset, src_grid_slice = self.src_grid.get_intersects(
sub_box,
keep_touches=False,
cascade=False,
optimized_bbox_subset=self.optimized_bbox_subset,
return_slice=True)
ocgis_lh(logger='grid_chunker',
msg='finished "self.src_grid.get_intersects"',
level=logging.DEBUG)
# Reload the data using a new source index distribution.
if hasattr(src_grid_subset,
'reduce_global') and src_grid_subset.cindex is not None:
# Only redistribute if we have one live rank.
if self.redistribute and len(
vm.get_live_ranks_from_object(src_grid_subset)) > 0:
ocgis_lh(logger='grid_chunker',
msg='starting redistribute',
level=logging.DEBUG)
topology = src_grid_subset.abstractions_available[
Topology.POLYGON]
cindex = topology.cindex
redist_dimname = self.src_grid.abstractions_available[
Topology.POLYGON].element_dim.name
if src_grid_subset.is_empty:
redist_dim = None
else:
redist_dim = topology.element_dim
redistribute_by_src_idx(cindex, redist_dimname, redist_dim)
ocgis_lh(logger='grid_chunker',
msg='finished redistribute',
level=logging.DEBUG)
with vm.scoped_by_emptyable('src_grid_subset', src_grid_subset):
if not vm.is_null:
if not self.allow_masked:
gmask = src_grid_subset.get_mask()
if gmask is not None and gmask.any():
raise ValueError(
'Masked values in source grid subset.')
if self.check_contains:
src_box = box(*src_grid_subset.extent_global)
if not does_contain(src_box, dst_box):
raise ValueError('Contains check failed.')
# Try to reduce the coordinates in the case of unstructured grid data.
if hasattr(src_grid_subset, 'reduce_global'
) and src_grid_subset.cindex is not None:
ocgis_lh(logger='grid_chunker',
msg='starting reduce_global',
level=logging.DEBUG)
src_grid_subset = src_grid_subset.reduce_global()
ocgis_lh(logger='grid_chunker',
msg='finished reduce_global',
level=logging.DEBUG)
else:
pass
# src_grid_subset = VariableCollection(is_empty=True)
if src_grid_subset.is_empty:
src_grid_slice = None
else:
src_grid_slice = {
src_grid_subset.dimensions[ii].name: src_grid_slice[ii]
for ii in range(src_grid_subset.ndim)
}
if yield_dst:
yld = (src_grid_subset, src_grid_slice, dst_grid_subset,
dst_slice)
else:
yld = src_grid_subset, src_grid_slice
yield yld
def global_grid_shape(grid):
with vm.scoped_by_emptyable('global grid shape', grid):
if not vm.is_null:
return grid.shape_global
def write_chunks(self):
"""
Write grid subsets to netCDF files using the provided filename templates. This will also generate ESMF
regridding weights for each subset if requested.
"""
src_filenames = []
dst_filenames = []
wgt_filenames = []
dst_slices = []
src_slices = []
index_path = self.create_full_path_from_template('index_file')
# nzeros = len(str(reduce(lambda x, y: x * y, self.nchunks_dst)))
ctr = 1
ocgis_lh(logger='grid_chunker', msg='starting self.iter_src_grid_subsets', level=logging.DEBUG)
for sub_src, src_slc, sub_dst, dst_slc in self.iter_src_grid_subsets(yield_dst=True):
ocgis_lh(logger='grid_chunker', msg='finished iteration {} for self.iter_src_grid_subsets'.format(ctr),
level=logging.DEBUG)
src_path = self.create_full_path_from_template('src_template', index=ctr)
dst_path = self.create_full_path_from_template('dst_template', index=ctr)
wgt_path = self.create_full_path_from_template('wgt_template', index=ctr)
src_filenames.append(os.path.split(src_path)[1])
dst_filenames.append(os.path.split(dst_path)[1])
wgt_filenames.append(wgt_path)
dst_slices.append(dst_slc)
src_slices.append(src_slc)
# Only write destinations if an iterator is not provided.
if self.iter_dst is None:
zip_args = [[sub_src, sub_dst], [src_path, dst_path]]
else:
zip_args = [[sub_src], [src_path]]
cc = 1
for target, path in zip(*zip_args):
with vm.scoped_by_emptyable('field.write' + str(cc), target):
if not vm.is_null:
ocgis_lh(logger='grid_chunker', msg='write_chunks:writing: {}'.format(path),
level=logging.DEBUG)
field = Field(grid=target)
field.write(path)
ocgis_lh(logger='grid_chunker', msg='write_chunks:finished writing: {}'.format(path),
level=logging.DEBUG)
cc += 1
# Increment the counter outside of the loop to avoid counting empty subsets.
ctr += 1
# Generate an ESMF weights file if requested and at least one rank has data on it.
if self.genweights and len(vm.get_live_ranks_from_object(sub_src)) > 0:
vm.barrier()
self.write_esmf_weights(src_path, dst_path, wgt_path, src_grid=sub_src, dst_grid=sub_dst)
vm.barrier()
# Global shapes require a VM global scope to collect.
src_global_shape = global_grid_shape(self.src_grid)
dst_global_shape = global_grid_shape(self.dst_grid)
# Gather and collapse source slices as some may be empty and we write on rank 0.
gathered_src_grid_slice = vm.gather(src_slices)
if vm.rank == 0:
len_src_slices = len(src_slices)
new_src_grid_slice = [None] * len_src_slices
for idx in range(len_src_slices):
for rank_src_grid_slice in gathered_src_grid_slice:
if rank_src_grid_slice[idx] is not None:
new_src_grid_slice[idx] = rank_src_grid_slice[idx]
break
src_slices = new_src_grid_slice
with vm.scoped('index write', [0]):
if not vm.is_null:
dim = Dimension('nfiles', len(src_filenames))
vname = ['source_filename', 'destination_filename', 'weights_filename']
values = [src_filenames, dst_filenames, wgt_filenames]
grid_chunker_destination = GridChunkerConstants.IndexFile.NAME_DESTINATION_VARIABLE
attrs = [{'esmf_role': 'grid_chunker_source'},
{'esmf_role': grid_chunker_destination},
{'esmf_role': 'grid_chunker_weights'}]
vc = VariableCollection()
grid_chunker_index = GridChunkerConstants.IndexFile.NAME_INDEX_VARIABLE
vidx = Variable(name=grid_chunker_index)
vidx.attrs['esmf_role'] = grid_chunker_index
vidx.attrs['grid_chunker_source'] = 'source_filename'
vidx.attrs[GridChunkerConstants.IndexFile.NAME_DESTINATION_VARIABLE] = 'destination_filename'
vidx.attrs['grid_chunker_weights'] = 'weights_filename'
vidx.attrs[GridChunkerConstants.IndexFile.NAME_SRC_GRID_SHAPE] = src_global_shape
vidx.attrs[GridChunkerConstants.IndexFile.NAME_DST_GRID_SHAPE] = dst_global_shape
vc.add_variable(vidx)
for idx in range(len(vname)):
v = Variable(name=vname[idx], dimensions=dim, dtype=str, value=values[idx], attrs=attrs[idx])
vc.add_variable(v)
bounds_dimension = Dimension(name='bounds', size=2)
# TODO: This needs to work with four dimensions.
# Source -----------------------------------------------------------------------------------------------
self.src_grid._gc_create_index_bounds_(RegriddingRole.SOURCE, vidx, vc, src_slices, dim,
bounds_dimension)
# Destination ------------------------------------------------------------------------------------------
self.dst_grid._gc_create_index_bounds_(RegriddingRole.DESTINATION, vidx, vc, dst_slices, dim,
bounds_dimension)
vc.write(index_path)
vm.barrier()
def iter_src_grid_subsets(self, yield_dst=False, yield_idx=None):
"""
Yield source grid subset using the extent of its associated destination grid subset.
:param bool yield_dst: If ``True``, yield the destination subset as well as the source grid subset.
:param int yield_idx: If a zero-based integer, only yield for this chunk index and skip everything else.
:rtype: tuple(:class:`ocgis.spatial.grid.AbstractGrid`, `slice-like`)
"""
if yield_dst:
yield_slice = True
else:
yield_slice = False
buffer_value = self.buffer_value
dst_grid_wrapped_state = self.dst_grid.wrapped_state
dst_grid_crs = self.dst_grid.crs
# Use a destination grid iterator if provided.
if self.iter_dst is not None:
iter_dst = self.iter_dst(self, yield_slice=yield_slice, yield_idx=yield_idx)
else:
iter_dst = self.iter_dst_grid_subsets(yield_slice=yield_slice, yield_idx=yield_idx)
# Loop over each destination grid subset.
ocgis_lh(logger='grid_chunker', msg='starting "for yld in iter_dst"', level=logging.DEBUG)
for iter_dst_ctr, yld in enumerate(iter_dst, start=1):
ocgis_lh(msg=["iter_dst_ctr", iter_dst_ctr], level=logging.DEBUG)
if yield_slice:
dst_grid_subset, dst_slice = yld
else:
dst_grid_subset = yld
# All masked destinations are very problematic for ESMF
with vm.scoped_by_emptyable('global mask', dst_grid_subset):
if not vm.is_null:
if dst_grid_subset.has_mask_global:
if dst_grid_subset.has_mask and dst_grid_subset.has_masked_values:
all_masked = dst_grid_subset.get_mask().all()
else:
all_masked = False
all_masked_gather = vm.gather(all_masked)
if vm.rank == 0:
if all(all_masked_gather):
exc = ValueError("Destination subset all masked")
try:
raise exc
finally:
vm.abort(exc=exc)
dst_box = None
with vm.scoped_by_emptyable('extent_global', dst_grid_subset):
if not vm.is_null:
# Use the extent of the polygon for determining the bounding box. This ensures conservative
# regridding will be fully mapped.
if isinstance(dst_grid_subset, AbstractGeometryCoordinates):
target_grid = dst_grid_subset.parent.grid
else:
target_grid = dst_grid_subset
# Try to reduce the coordinates in the case of unstructured grid data.
if hasattr(target_grid, 'reduce_global') and Topology.POLYGON in target_grid.abstractions_available:
ocgis_lh(logger='grid_chunker', msg='starting reduce_global for dst_grid_subset',
level=logging.DEBUG)
target_grid = target_grid.reduce_global()
ocgis_lh(logger='grid_chunker', msg='finished reduce_global for dst_grid_subset',
level=logging.DEBUG)
extent_global = target_grid.parent.attrs.get('extent_global')
if extent_global is None:
with grid_abstraction_scope(target_grid, Topology.POLYGON):
extent_global = target_grid.extent_global
if self.check_contains:
dst_box = box(*target_grid.extent_global)
sub_box = box(*extent_global)
if buffer_value is not None:
# Use the envelope! A buffer returns "fancy" borders. We just want to expand the bounding box.
sub_box = sub_box.buffer(buffer_value).envelope
ocgis_lh(msg=str(sub_box.bounds), level=logging.DEBUG, logger='grid_chunker')
else:
sub_box, dst_box = [None, None]
live_ranks = vm.get_live_ranks_from_object(dst_grid_subset)
sub_box = vm.bcast(sub_box, root=live_ranks[0])
if self.check_contains:
dst_box = vm.bcast(dst_box, root=live_ranks[0])
sub_box = GeometryVariable.from_shapely(sub_box, is_bbox=True, wrapped_state=dst_grid_wrapped_state,
crs=dst_grid_crs)
ocgis_lh(logger='grid_chunker', msg='starting "self.src_grid.get_intersects"', level=logging.DEBUG)
src_grid_subset, src_grid_slice = self.src_grid.get_intersects(sub_box, keep_touches=False, cascade=False,
optimized_bbox_subset=self.optimized_bbox_subset,
return_slice=True)
ocgis_lh(logger='grid_chunker', msg='finished "self.src_grid.get_intersects"', level=logging.DEBUG)
# Reload the data using a new source index distribution.
if hasattr(src_grid_subset, 'reduce_global') and src_grid_subset.cindex is not None:
# Only redistribute if we have one live rank.
if self.redistribute and len(vm.get_live_ranks_from_object(src_grid_subset)) > 0:
ocgis_lh(logger='grid_chunker', msg='starting redistribute', level=logging.DEBUG)
topology = src_grid_subset.abstractions_available[Topology.POLYGON]
cindex = topology.cindex
redist_dimname = self.src_grid.abstractions_available[Topology.POLYGON].element_dim.name
if src_grid_subset.is_empty:
redist_dim = None
else:
redist_dim = topology.element_dim
redistribute_by_src_idx(cindex, redist_dimname, redist_dim)
ocgis_lh(logger='grid_chunker', msg='finished redistribute', level=logging.DEBUG)
with vm.scoped_by_emptyable('src_grid_subset', src_grid_subset):
if not vm.is_null:
if not self.allow_masked:
gmask = src_grid_subset.get_mask()
if gmask is not None and gmask.any():
raise ValueError('Masked values in source grid subset.')
if self.check_contains:
src_box = box(*src_grid_subset.extent_global)
if not does_contain(src_box, dst_box):
raise ValueError('Contains check failed.')
# Try to reduce the coordinates in the case of unstructured grid data.
if hasattr(src_grid_subset, 'reduce_global') and src_grid_subset.cindex is not None:
ocgis_lh(logger='grid_chunker', msg='starting reduce_global', level=logging.DEBUG)
src_grid_subset = src_grid_subset.reduce_global()
ocgis_lh(logger='grid_chunker', msg='finished reduce_global', level=logging.DEBUG)
else:
pass
# src_grid_subset = VariableCollection(is_empty=True)
if src_grid_subset.is_empty:
src_grid_slice = None
else:
src_grid_slice = {src_grid_subset.dimensions[ii].name: src_grid_slice[ii] for ii in
range(src_grid_subset.ndim)}
if yield_dst:
yld = (src_grid_subset, src_grid_slice, dst_grid_subset, dst_slice)
else:
yld = src_grid_subset, src_grid_slice
yield yld