def test_9_TestView(self):
"""
Load some of the dataset again; this time with an offset view.
Note: The original blockwise fileset must be closed before this test starts.
"""
# Create a copy of the original description, but specify a translated (and smaller) view
desc = BlockwiseFileset.readDescription(self.description_path)
desc.view_origin = [0, 300, 200, 100, 0]
desc.view_shape = [1, 50, 50, 50, 1]
offsetConfigPath = self.description_path + "_offset"
BlockwiseFileset.writeDescription(offsetConfigPath, desc)
# Open the fileset using the special description file
bfs = BlockwiseFileset(offsetConfigPath, "r")
try:
assert (bfs.description.view_origin == desc.view_origin).all()
assert (bfs.description.view_shape == desc.view_shape).all()
# Read some data
logger.debug("Reading data...")
disk_slicing = numpy.s_[:, 300:350, 200:250, 100:150, :]
view_slicing = numpy.s_[:, 0:50, 0:50, 0:50, :]
roi = sliceToRoi(view_slicing, self.dataShape)
roiShape = roi[1] - roi[0]
read_data = numpy.zeros(tuple(roiShape), dtype=numpy.uint8)
bfs.readData(roi, read_data)
# The data we read should match the correct part of the original dataset.
logger.debug("Checking data...")
assert self.data[disk_slicing].shape == read_data.shape
assert (self.data[disk_slicing] == read_data).all(), "Data didn't match."
finally:
bfs.close()
class OpBlockwiseFilesetReader(Operator):
"""
Adapter that provides an operator interface to the BlockwiseFileset class for reading ONLY.
"""
name = "OpBlockwiseFilesetReader"
DescriptionFilePath = InputSlot(stype="filestring")
Output = OutputSlot()
class MissingDatasetError(Exception):
pass
def __init__(self, *args, **kwargs):
super(OpBlockwiseFilesetReader, self).__init__(*args, **kwargs)
self._blockwiseFileset = None
self._opDummyData = OpDummyData(parent=self)
def setupOutputs(self):
if not os.path.exists(self.DescriptionFilePath.value):
raise OpBlockwiseFilesetReader.MissingDatasetError(
"Dataset description not found: {}".format(
self.DescriptionFilePath.value))
# Load up the class that does the real work
self._blockwiseFileset = BlockwiseFileset(
self.DescriptionFilePath.value)
# Check for errors in the description file
descriptionFields = self._blockwiseFileset.description
axes = descriptionFields.axes
assert False not in [
a in "txyzc" for a in axes
], "Unknown axis type. Known axes: txyzc Your axes:".format(axes)
self.Output.meta.shape = tuple(descriptionFields.view_shape)
self.Output.meta.dtype = descriptionFields.dtype
self.Output.meta.axistags = vigra.defaultAxistags(
str(descriptionFields.axes))
drange = descriptionFields.drange
if drange is not None:
self.Output.meta.drange = drange
def execute(self, slot, subindex, roi, result):
assert slot == self.Output, "Unknown output slot"
try:
self._blockwiseFileset.readData((roi.start, roi.stop), result)
except BlockwiseFileset.BlockNotReadyError:
result[:] = self._opDummyData.execute(slot, subindex, roi, result)
return result
def propagateDirty(self, slot, subindex, roi):
assert slot == self.DescriptionFilePath, "Unknown input slot."
self.Output.setDirty(slice(None))
def cleanUp(self):
if self._blockwiseFileset is not None:
self._blockwiseFileset.close()
super(OpBlockwiseFilesetReader, self).cleanUp()
class OpBlockwiseFilesetReader(Operator):
"""
Adapter that provides an operator interface to the BlockwiseFileset class for reading ONLY.
"""
name = "OpBlockwiseFilesetReader"
DescriptionFilePath = InputSlot(stype="filestring")
Output = OutputSlot()
class MissingDatasetError(Exception):
pass
def __init__(self, *args, **kwargs):
super(OpBlockwiseFilesetReader, self).__init__(*args, **kwargs)
self._blockwiseFileset = None
self._opDummyData = OpDummyData(parent=self)
def setupOutputs(self):
if not os.path.exists(self.DescriptionFilePath.value):
raise OpBlockwiseFilesetReader.MissingDatasetError(
"Dataset description not found: {}".format(self.DescriptionFilePath.value)
)
# Load up the class that does the real work
self._blockwiseFileset = BlockwiseFileset(self.DescriptionFilePath.value)
# Check for errors in the description file
descriptionFields = self._blockwiseFileset.description
axes = descriptionFields.axes
assert False not in [a in "txyzc" for a in axes], "Unknown axis type. Known axes: txyzc Your axes:".format(
axes
)
self.Output.meta.shape = tuple(descriptionFields.view_shape)
self.Output.meta.dtype = descriptionFields.dtype
self.Output.meta.axistags = vigra.defaultAxistags(str(descriptionFields.axes))
drange = descriptionFields.drange
if drange is not None:
self.Output.meta.drange = drange
def execute(self, slot, subindex, roi, result):
assert slot == self.Output, "Unknown output slot"
try:
self._blockwiseFileset.readData((roi.start, roi.stop), result)
except BlockwiseFileset.BlockNotReadyError:
result[:] = self._opDummyData.execute(slot, subindex, roi, result)
return result
def propagateDirty(self, slot, subindex, roi):
assert slot == self.DescriptionFilePath, "Unknown input slot."
self.Output.setDirty(slice(None))
def cleanUp(self):
if self._blockwiseFileset is not None:
self._blockwiseFileset.close()
super(OpBlockwiseFilesetReader, self).cleanUp()
def setUp(self):
"""
Create a blockwise fileset to test with.
"""
if platform.system() == 'Windows':
# On windows, there are errors, and we make no attempt to solve them (at the moment).
raise nose.SkipTest
try:
BlockwiseFileset._prepare_system()
except ValueError:
# If the system isn't configured to allow lots of open files, we can't run this test.
raise nose.SkipTest
testConfig = \
"""
{
"_schema_name" : "blockwise-fileset-description",
"_schema_version" : 1.0,
"name" : "synapse_small",
"format" : "hdf5",
"axes" : "txyzc",
"shape" : [1,400,400,100,1],
"dtype" : "numpy.uint8",
"block_shape" : [1, 50, 50, 50, 100],
"block_file_name_format" : "cube{roiString}.h5/volume/data"
}
"""
self.tempDir = tempfile.mkdtemp()
self.configpath = os.path.join(self.tempDir, "config.json")
logger.debug("Loading config file...")
with open(self.configpath, 'w') as f:
f.write(testConfig)
logger.debug("Creating random test data...")
bfs = BlockwiseFileset(self.configpath, 'a')
dataShape = tuple(bfs.description.shape)
self.data = numpy.random.randint(255,
size=dataShape).astype(numpy.uint8)
logger.debug("Writing test data...")
datasetRoi = ([0, 0, 0, 0, 0], dataShape)
bfs.writeData(datasetRoi, self.data)
block_starts = getIntersectingBlocks(bfs.description.block_shape,
datasetRoi)
for block_start in block_starts:
bfs.setBlockStatus(block_start, BlockwiseFileset.BLOCK_AVAILABLE)
bfs.close()
def setUp(self):
"""
Create a blockwise fileset to test with.
"""
if platform.system() == 'Windows':
# On windows, there are errors, and we make no attempt to solve them (at the moment).
raise nose.SkipTest
try:
BlockwiseFileset._prepare_system()
except ValueError:
# If the system isn't configured to allow lots of open files, we can't run this test.
raise nose.SkipTest
testConfig = \
"""
{
"_schema_name" : "blockwise-fileset-description",
"_schema_version" : 1.0,
"name" : "synapse_small",
"format" : "hdf5",
"axes" : "txyzc",
"shape" : [1,400,400,100,1],
"dtype" : "numpy.uint8",
"block_shape" : [1, 50, 50, 50, 100],
"block_file_name_format" : "cube{roiString}.h5/volume/data"
}
"""
self.tempDir = tempfile.mkdtemp()
self.configpath = os.path.join(self.tempDir, "config.json")
logger.debug( "Loading config file..." )
with open(self.configpath, 'w') as f:
f.write(testConfig)
logger.debug( "Creating random test data..." )
bfs = BlockwiseFileset( self.configpath, 'a' )
dataShape = tuple(bfs.description.shape)
self.data = numpy.random.randint( 255, size=dataShape ).astype(numpy.uint8)
logger.debug( "Writing test data..." )
datasetRoi = ([0,0,0,0,0], dataShape)
bfs.writeData( datasetRoi, self.data )
block_starts = getIntersectingBlocks(bfs.description.block_shape, datasetRoi)
for block_start in block_starts:
bfs.setBlockStatus(block_start, BlockwiseFileset.BLOCK_AVAILABLE)
bfs.close()
def test_9_TestView(self):
"""
Load some of the dataset again; this time with an offset view.
Note: The original blockwise fileset must be closed before this test starts.
"""
# Create a copy of the original description, but specify a translated (and smaller) view
desc = BlockwiseFileset.readDescription(self.description_path)
desc.view_origin = [0, 300, 200, 100, 0]
desc.view_shape = [1, 50, 50, 50, 1]
offsetConfigPath = self.description_path + "_offset"
BlockwiseFileset.writeDescription(offsetConfigPath, desc)
# Open the fileset using the special description file
bfs = BlockwiseFileset(offsetConfigPath, "r")
try:
assert (bfs.description.view_origin == desc.view_origin).all()
assert (bfs.description.view_shape == desc.view_shape).all()
# Read some data
logger.debug("Reading data...")
disk_slicing = numpy.s_[:, 300:350, 200:250, 100:150, :]
view_slicing = numpy.s_[:, 0:50, 0:50, 0:50, :]
roi = sliceToRoi(view_slicing, self.dataShape)
roiShape = roi[1] - roi[0]
read_data = numpy.zeros(tuple(roiShape), dtype=numpy.uint8)
bfs.readData(roi, read_data)
# The data we read should match the correct part of the original dataset.
logger.debug("Checking data...")
assert self.data[disk_slicing].shape == read_data.shape
assert (self.data[disk_slicing] == read_data
).all(), "Data didn't match."
finally:
bfs.close()
class OpTaskWorker(Operator):
Input = InputSlot()
RoiString = InputSlot(stype="string")
TaskName = InputSlot(stype="string")
ConfigFilePath = InputSlot(stype="filestring")
OutputFilesetDescription = InputSlot(stype="filestring")
ReturnCode = OutputSlot()
def __init__(self, *args, **kwargs):
super(OpTaskWorker, self).__init__(*args, **kwargs)
self.progressSignal = OrderedSignal()
self._primaryBlockwiseFileset = None
def setupOutputs(self):
self.ReturnCode.meta.dtype = bool
self.ReturnCode.meta.shape = (1,)
self._closeFiles()
self._primaryBlockwiseFileset = BlockwiseFileset(self.OutputFilesetDescription.value, "a")
def cleanUp(self):
self._closeFiles()
super(OpTaskWorker, self).cleanUp()
def _closeFiles(self):
if self._primaryBlockwiseFileset is not None:
self._primaryBlockwiseFileset.close()
self._primaryBlockwiseFileset = None
def execute(self, slot, subindex, ignored_roi, result):
configFilePath = self.ConfigFilePath.value
config = parseClusterConfigFile(configFilePath)
blockwiseFileset = self._primaryBlockwiseFileset
# Check axis compatibility
inputAxes = list(self.Input.meta.getTaggedShape().keys())
outputAxes = list(blockwiseFileset.description.axes)
assert set(inputAxes) == set(
outputAxes
), "Output dataset has the wrong set of axes. Input axes: {}, Output axes: {}".format(
"".join(inputAxes), "".join(outputAxes)
)
roiString = self.RoiString.value
roi = Roi.loads(roiString)
if len(roi.start) != len(self.Input.meta.shape):
assert (
False
), "Task roi: {} is not valid for this input. Did the master launch this task correctly?".format(roiString)
logger.info("Executing for roi: {}".format(roi))
if config.use_node_local_scratch:
assert False, "FIXME."
assert (
blockwiseFileset.getEntireBlockRoi(roi.start)[1] == roi.stop
).all(), "Each task must execute exactly one full block. ({},{}) is not a valid block roi.".format(
roi.start, roi.stop
)
assert self.Input.ready()
with Timer() as computeTimer:
# Stream the data out to disk.
request_blockshape = (
self._primaryBlockwiseFileset.description.sub_block_shape
) # Could be None. That's okay.
streamer = BigRequestStreamer(self.Input, (roi.start, roi.stop), request_blockshape)
streamer.progressSignal.subscribe(self.progressSignal)
streamer.resultSignal.subscribe(self._handlePrimaryResultBlock)
streamer.execute()
# Now the block is ready. Update the status.
blockwiseFileset.setBlockStatus(roi.start, BlockwiseFileset.BLOCK_AVAILABLE)
logger.info("Finished task in {} seconds".format(computeTimer.seconds()))
result[0] = True
return result
def propagateDirty(self, slot, subindex, roi):
self.ReturnCode.setDirty(slice(None))
def _handlePrimaryResultBlock(self, roi, result):
# First write the primary
self._primaryBlockwiseFileset.writeData(roi, result)
# Ask the workflow if there is any special post-processing to do...
self.get_workflow().postprocessClusterSubResult(roi, result, self._primaryBlockwiseFileset)
def get_workflow(self):
op = self
while not isinstance(op, Workflow):
op = op.parent
return op
class OpTaskWorker(Operator):
Input = InputSlot()
RoiString = InputSlot(stype='string')
TaskName = InputSlot(stype='string')
ConfigFilePath = InputSlot(stype='filestring')
OutputFilesetDescription = InputSlot(stype='filestring')
ReturnCode = OutputSlot()
def __init__(self, *args, **kwargs):
super( OpTaskWorker, self ).__init__( *args, **kwargs )
self.progressSignal = OrderedSignal()
self._primaryBlockwiseFileset = None
def setupOutputs(self):
self.ReturnCode.meta.dtype = bool
self.ReturnCode.meta.shape = (1,)
self._closeFiles()
self._primaryBlockwiseFileset = BlockwiseFileset( self.OutputFilesetDescription.value, 'a' )
def cleanUp(self):
self._closeFiles()
super( OpTaskWorker, self ).cleanUp()
def _closeFiles(self):
if self._primaryBlockwiseFileset is not None:
self._primaryBlockwiseFileset.close()
self._primaryBlockwiseFileset = None
def execute(self, slot, subindex, ignored_roi, result):
configFilePath = self.ConfigFilePath.value
config = parseClusterConfigFile( configFilePath )
blockwiseFileset = self._primaryBlockwiseFileset
# Check axis compatibility
inputAxes = list(self.Input.meta.getTaggedShape().keys())
outputAxes = list(blockwiseFileset.description.axes)
assert set(inputAxes) == set(outputAxes), \
"Output dataset has the wrong set of axes. Input axes: {}, Output axes: {}".format( "".join(inputAxes), "".join(outputAxes) )
roiString = self.RoiString.value
roi = Roi.loads(roiString)
if len( roi.start ) != len( self.Input.meta.shape ):
assert False, "Task roi: {} is not valid for this input. Did the master launch this task correctly?".format( roiString )
logger.info( "Executing for roi: {}".format(roi) )
if config.use_node_local_scratch:
assert False, "FIXME."
assert (blockwiseFileset.getEntireBlockRoi( roi.start )[1] == roi.stop).all(), "Each task must execute exactly one full block. ({},{}) is not a valid block roi.".format( roi.start, roi.stop )
assert self.Input.ready()
with Timer() as computeTimer:
# Stream the data out to disk.
request_blockshape = self._primaryBlockwiseFileset.description.sub_block_shape # Could be None. That's okay.
streamer = BigRequestStreamer(self.Input, (roi.start, roi.stop), request_blockshape )
streamer.progressSignal.subscribe( self.progressSignal )
streamer.resultSignal.subscribe( self._handlePrimaryResultBlock )
streamer.execute()
# Now the block is ready. Update the status.
blockwiseFileset.setBlockStatus( roi.start, BlockwiseFileset.BLOCK_AVAILABLE )
logger.info( "Finished task in {} seconds".format( computeTimer.seconds() ) )
result[0] = True
return result
def propagateDirty(self, slot, subindex, roi):
self.ReturnCode.setDirty( slice(None) )
def _handlePrimaryResultBlock(self, roi, result):
# First write the primary
self._primaryBlockwiseFileset.writeData(roi, result)
# Ask the workflow if there is any special post-processing to do...
self.get_workflow().postprocessClusterSubResult(roi, result, self._primaryBlockwiseFileset)
def get_workflow(self):
op = self
while not isinstance(op, Workflow):
op = op.parent
return op
