def test_toBlocksBySlices(self):
narys = 3
arys, sh, sz = _generate_test_arrays(narys)
imagedata = ImagesLoader(self.sc).fromArrays(arys)
test_params = [
(1, 1, 1), (1, 1, 2), (1, 1, 3), (1, 2, 1), (1, 2, 2), (1, 2, 3),
(1, 3, 1), (1, 3, 2), (1, 3, 3),
(2, 1, 1), (2, 1, 2), (2, 1, 3), (2, 2, 1), (2, 2, 2), (2, 2, 3),
(2, 3, 1), (2, 3, 2), (2, 3, 3)]
for bpd in test_params:
blocks = imagedata._toBlocksBySplits(bpd).collect()
expectednuniquekeys = reduce(mul, bpd)
expectedvalsperkey = narys
keystocounts = Counter([kv[0] for kv in blocks])
assert_equals(expectednuniquekeys, len(keystocounts))
assert_equals([expectedvalsperkey] * expectednuniquekeys, keystocounts.values())
gatheredary = None
for _, block in blocks:
if gatheredary is None:
gatheredary = zeros(block.origshape, dtype='int16')
gatheredary[block.origslices] = block.values
for i in xrange(narys):
assert_true(array_equal(arys[i], gatheredary[i]))
def test_toSeriesWithInefficientSplitAndSortedPack(self):
ary = arange(8, dtype=dtypeFunc("int16")).reshape((4, 2))
image = ImagesLoader(self.sc).fromArrays(ary)
series = image.toBlocks((2, 1), units="s").toSeries()
seriesVals = series.collect()
seriesAry = series.pack(sorting=True)
# check ordering of keys
assert_equals((0, 0), seriesVals[0][0]) # first key
assert_equals((1, 0), seriesVals[1][0]) # second key
assert_equals((0, 1), seriesVals[2][0])
assert_equals((1, 1), seriesVals[3][0])
# end of first block
# beginning of second block
assert_equals((2, 0), seriesVals[4][0])
assert_equals((3, 0), seriesVals[5][0])
assert_equals((2, 1), seriesVals[6][0])
assert_equals((3, 1), seriesVals[7][0])
# check dimensions tuple matches numpy shape
assert_equals(ary.shape, series.dims.count)
# check that values are in expected order
collectedVals = array([kv[1] for kv in seriesVals], dtype=dtypeFunc("int16")).ravel()
assert_true(array_equal(ary[:2, :].ravel(order="F"), collectedVals[:4])) # first block
assert_true(array_equal(ary[2:4, :].ravel(order="F"), collectedVals[4:])) # second block
# check that packing returns original array (after sort)
assert_true(array_equal(ary, seriesAry))
def test_castToFloat(self):
arys, shape, size = _generateTestArrays(2, "uint8")
imageData = ImagesLoader(self.sc).fromArrays(arys)
catData = imageData.astype("smallfloat")
assert_equals("float16", str(catData.dtype))
assert_equals("float16", str(catData.first()[1].dtype))
def test_fromMultiTimepointStacks(self):
ary = arange(16, dtype=dtypeFunc('uint8')).reshape((4, 2, 2))
ary2 = arange(16, 32, dtype=dtypeFunc('uint8')).reshape((4, 2, 2))
ary.tofile(os.path.join(self.outputdir, "test01.stack"))
ary2.tofile(os.path.join(self.outputdir, "test02.stack"))
image = ImagesLoader(self.sc).fromStack(self.outputdir, dtype="uint8", dims=(2, 2, 4), nplanes=2)
collectedImage = image.collect()
# we don't expect to have nrecords cached, since we get an unknown number of images per file
assert_true(image._nrecords is None)
assert_equals(4, image.nrecords)
assert_equals(4, len(collectedImage))
# check keys:
assert_equals(0, collectedImage[0][0])
assert_equals(1, collectedImage[1][0])
assert_equals(2, collectedImage[2][0])
assert_equals(3, collectedImage[3][0])
# check values:
assert_true(array_equal(ary[:2].T, collectedImage[0][1]))
assert_true(array_equal(ary[2:].T, collectedImage[1][1]))
assert_true(array_equal(ary2[:2].T, collectedImage[2][1]))
assert_true(array_equal(ary2[2:].T, collectedImage[3][1]))
# 3 planes does not divide 4
assert_raises(ValueError, ImagesLoader(self.sc).fromStack, self.outputdir, dtype="uint8",
dims=(2, 2, 4), nplanes=3)
def loadImagesFromArray(self, values, npartitions=None):
"""
Load Images data from a local array
Parameters
----------
values : list or ndarray
A list of 2d or 3d numpy arrays,
or a single 3d or 4d numpy array
npartitions : position int, optional, default = None
Number of partitions for RDD, if unspecified will use
default parallelism.
"""
from numpy import ndarray, asarray
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if isinstance(values, list):
values = asarray(values)
if isinstance(values, ndarray) and values.ndim > 2:
values = list(values)
if not npartitions:
npartitions = self._sc.defaultParallelism
return loader.fromArrays(values, npartitions=npartitions)
def _run_tst_toSeriesWithSplitsAndPack(self, strategy):
ary = arange(8, dtype=dtypeFunc('int16')).reshape((4, 2))
image = ImagesLoader(self.sc).fromArrays(ary)
series = image.toBlocks(strategy).toSeries()
seriesVals = series.collect()
seriesAry = series.pack()
# check ordering of keys
assert_equals((0, 0), seriesVals[0][0]) # first key
assert_equals((1, 0), seriesVals[1][0]) # second key
assert_equals((2, 0), seriesVals[2][0])
assert_equals((3, 0), seriesVals[3][0])
assert_equals((0, 1), seriesVals[4][0])
assert_equals((1, 1), seriesVals[5][0])
assert_equals((2, 1), seriesVals[6][0])
assert_equals((3, 1), seriesVals[7][0])
# check dimensions tuple matches numpy shape
assert_equals(ary.shape, series.dims.count)
# check that values are in Fortran-convention order
collectedVals = array([kv[1] for kv in seriesVals], dtype=dtypeFunc('int16')).ravel()
assert_true(array_equal(ary.ravel(order='F'), collectedVals))
# check that packing returns original array
assert_true(array_equal(ary, seriesAry))
def test_toSeriesBySlices(self):
narys = 3
arys, sh, sz = _generateTestArrays(narys)
imageData = ImagesLoader(self.sc).fromArrays(arys)
imageData.cache()
testParams = [
(1, 1, 1),
(1, 1, 2),
(1, 1, 3),
(1, 2, 1),
(1, 2, 2),
(1, 2, 3),
(1, 3, 1),
(1, 3, 2),
(1, 3, 3),
(2, 1, 1),
(2, 1, 2),
(2, 1, 3),
(2, 2, 1),
(2, 2, 2),
(2, 2, 3),
(2, 3, 1),
(2, 3, 2),
(2, 3, 3),
]
for bpd in testParams:
series = imageData.toBlocks(bpd, units="s").toSeries().collect()
self.evaluateSeries(arys, series, sz)
def setUp(self):
super(TestBlockKeys, self).setUp()
shape = (30, 30)
arys = [ones(shape) for _ in range(0, 3)]
data = ImagesLoader(self.sc).fromArrays(arys)
self.blocks = data.toBlocks(size=(10, 10)).collect()
self.keys = [k for k, v in self.blocks]
def test_toSeriesWithPack(self):
ary = arange(8, dtype=dtypeFunc("int16")).reshape((2, 4))
image = ImagesLoader(self.sc).fromArrays(ary)
series = image.toBlocks("150M").toSeries()
seriesVals = series.collect()
seriesAry = series.pack()
seriesAry_xpose = series.pack(transpose=True)
# check ordering of keys
assert_equals((0, 0), seriesVals[0][0]) # first key
assert_equals((1, 0), seriesVals[1][0]) # second key
assert_equals((0, 1), seriesVals[2][0])
assert_equals((1, 1), seriesVals[3][0])
assert_equals((0, 2), seriesVals[4][0])
assert_equals((1, 2), seriesVals[5][0])
assert_equals((0, 3), seriesVals[6][0])
assert_equals((1, 3), seriesVals[7][0])
# check dimensions tuple matches numpy shape
assert_equals(image.dims.count, series.dims.count)
assert_equals(ary.shape, series.dims.count)
# check that values are in Fortran-convention order
collectedVals = array([kv[1] for kv in seriesVals], dtype=dtypeFunc("int16")).ravel()
assert_true(array_equal(ary.ravel(order="F"), collectedVals))
# check that packing returns original array
assert_true(array_equal(ary, seriesAry))
assert_true(array_equal(ary.T, seriesAry_xpose))
def test_fromStackToSeriesWithPack(self):
ary = arange(8, dtype=dtypeFunc("int16")).reshape((2, 4))
filename = os.path.join(self.outputdir, "test.stack")
ary.tofile(filename)
image = ImagesLoader(self.sc).fromStack(filename, dims=(4, 2))
strategy = SimpleBlockingStrategy.generateFromBlockSize(image, "150M")
series = image.toBlocks(strategy).toSeries()
seriesVals = series.collect()
seriesAry = series.pack()
# check ordering of keys
assert_equals((0, 0), seriesVals[0][0]) # first key
assert_equals((1, 0), seriesVals[1][0]) # second key
assert_equals((2, 0), seriesVals[2][0])
assert_equals((3, 0), seriesVals[3][0])
assert_equals((0, 1), seriesVals[4][0])
assert_equals((1, 1), seriesVals[5][0])
assert_equals((2, 1), seriesVals[6][0])
assert_equals((3, 1), seriesVals[7][0])
# check dimensions tuple is reversed from numpy shape
assert_equals(ary.shape[::-1], series.dims.count)
# check that values are in original order
collectedVals = array([kv[1] for kv in seriesVals], dtype=dtypeFunc("int16")).ravel()
assert_true(array_equal(ary.ravel(), collectedVals))
# check that packing returns transpose of original array
assert_true(array_equal(ary.T, seriesAry))
def test_min(self):
from numpy import minimum
arys, shape, size = _generateTestArrays(2, "uint8")
imageData = ImagesLoader(self.sc).fromArrays(arys)
minVal = imageData.min()
assert_true(array_equal(reduce(minimum, arys), minVal))
def test_castToFloat(self):
arys, shape, size = _generate_test_arrays(2, 'uint8')
imagedata = ImagesLoader(self.sc).fromArrays(arys)
castdata = imagedata.astype("smallfloat")
assert_equals('float16', str(castdata.dtype))
assert_equals('float16', str(castdata.first()[1].dtype))
def test_mean(self):
from test_utils import elementwiseMean
arys, shape, size = _generateTestArrays(2, 'uint8')
imageData = ImagesLoader(self.sc).fromArrays(arys)
meanVal = imageData.mean()
expected = elementwiseMean(arys).astype('float16')
assert_true(allclose(expected, meanVal))
assert_equals('float64', str(meanVal.dtype))
def _run_tstSaveAsBinarySeries(self, testIdx, narys_, valDtype, groupingDim_):
"""Pseudo-parameterized test fixture, allows reusing existing spark context
"""
paramStr = "(groupingdim=%d, valuedtype='%s')" % (groupingDim_, valDtype)
arys, aryShape, arySize = _generateTestArrays(narys_, dtype_=valDtype)
dims = aryShape[:]
outdir = os.path.join(self.outputdir, "anotherdir%02d" % testIdx)
images = ImagesLoader(self.sc).fromArrays(arys)
slicesPerDim = [1]*arys[0].ndim
slicesPerDim[groupingDim_] = arys[0].shape[groupingDim_]
images.toBlocks(slicesPerDim, units="splits").saveAsBinarySeries(outdir)
ndims = len(aryShape)
# prevent padding to 4-byte boundaries: "=" specifies no alignment
unpacker = struct.Struct('=' + 'h'*ndims + dtypeFunc(valDtype).char*narys_)
def calcExpectedNKeys():
tmpShape = list(dims[:])
del tmpShape[groupingDim_]
return prod(tmpShape)
expectedNKeys = calcExpectedNKeys()
def byrec(f_, unpacker_, nkeys_):
rec = True
while rec:
rec = f_.read(unpacker_.size)
if rec:
allRecVals = unpacker_.unpack(rec)
yield allRecVals[:nkeys_], allRecVals[nkeys_:]
outFilenames = glob.glob(os.path.join(outdir, "*.bin"))
assert_equals(dims[groupingDim_], len(outFilenames))
for outFilename in outFilenames:
with open(outFilename, 'rb') as f:
nkeys = 0
for keys, vals in byrec(f, unpacker, ndims):
nkeys += 1
assert_equals(narys_, len(vals))
for valIdx, val in enumerate(vals):
assert_equals(arys[valIdx][keys], val, "Expected %g, got %g, for test %d %s" %
(arys[valIdx][keys], val, testIdx, paramStr))
assert_equals(expectedNKeys, nkeys)
confName = os.path.join(outdir, "conf.json")
assert_true(os.path.isfile(confName))
with open(os.path.join(outdir, "conf.json"), 'r') as fconf:
import json
conf = json.load(fconf)
assert_equals(outdir, conf['input'])
assert_equals(len(aryShape), conf['nkeys'])
assert_equals(narys_, conf['nvalues'])
assert_equals(valDtype, conf['valuetype'])
assert_equals('int16', conf['keytype'])
assert_true(os.path.isfile(os.path.join(outdir, 'SUCCESS')))
def test_stdev(self):
from test_utils import elementwiseStdev
arys, shape, size = _generateTestArrays(2, 'uint8')
imageData = ImagesLoader(self.sc).fromArrays(arys)
stdval = imageData.stdev()
expected = elementwiseStdev([ary.astype('float16') for ary in arys])
assert_true(allclose(expected, stdval))
assert_equals('float64', str(stdval.dtype))
def test_toTimeSeries(self):
# create 3 arrays of 4x3x3 images (C-order), containing sequential integers
narys = 3
arys, sh, sz = _generateTestArrays(narys)
imageData = ImagesLoader(self.sc).fromArrays(arys)
series = imageData.toTimeSeries()
assert isinstance(series, TimeSeries)
def test_toSeries(self):
# create 3 arrays of 4x3x3 images (C-order), containing sequential integers
narys = 3
arys, sh, sz = _generate_test_arrays(narys)
imagedata = ImagesLoader(self.sc).fromArrays(arys)
series = imagedata.toSeries(groupingDim=0).collect()
self.evaluate_series(arys, series, sz)
def test_toSeries(self):
# create 3 arrays of 4x3x3 images (C-order), containing sequential integers
narys = 3
arys, sh, sz = _generateTestArrays(narys)
imageData = ImagesLoader(self.sc).fromArrays(arys)
series = imageData.toBlocks((4, 1, 1), units="s").toSeries().collect()
self.evaluateSeries(arys, series, sz)
def test_variance(self):
from test_utils import elementwise_var
arys, shape, size = _generate_test_arrays(2, 'uint8')
imagedata = ImagesLoader(self.sc).fromArrays(arys)
varval = imagedata.variance()
expected = elementwise_var([ary.astype('float16') for ary in arys])
assert_true(allclose(expected, varval))
assert_equals('float16', str(varval.dtype))
def test_fromArrays(self):
ary = arange(8, dtype=dtypeFunc('int16')).reshape((2, 4))
image = ImagesLoader(self.sc).fromArrays(ary)
collectedImage = image.collect()
assert_equals(1, len(collectedImage))
assert_equals(ary.shape, image.dims.count)
assert_equals(0, collectedImage[0][0]) # check key
assert_true(array_equal(ary, collectedImage[0][1])) # check value
def _run_tst_roundtripThroughBlocks(self, strategy):
imagepath = findSourceTreeDir("utils/data/fish/images")
images = ImagesLoader(self.sc).fromTif(imagepath)
blockedimages = images.toBlocks(strategy)
recombinedimages = blockedimages.toImages()
collectedimages = images.collect()
roundtrippedimages = recombinedimages.collect()
for orig, roundtripped in zip(collectedimages, roundtrippedimages):
assert_true(array_equal(orig[1], roundtripped[1]))
def test_sum(self):
from numpy import add
arys, shape, size = _generateTestArrays(2, 'uint8')
imageData = ImagesLoader(self.sc).fromArrays(arys)
sumVal = imageData.sum(dtype='uint32')
arys = [ary.astype('uint32') for ary in arys]
expected = reduce(add, arys)
assert_true(array_equal(expected, sumVal))
assert_equals('uint32', str(sumVal.dtype))
def test_variance(self):
from test_utils import elementwiseVar
arys, shape, size = _generateTestArrays(2, "uint8")
imageData = ImagesLoader(self.sc).fromArrays(arys)
varVal = imageData.variance()
expected = elementwiseVar([ary.astype("float16") for ary in arys])
assert_true(allclose(expected, varVal))
assert_equals("float64", str(varVal.dtype))
def loadImagesOCP(
self, bucketName, resolution, server="ocp.me", startIdx=None, stopIdx=None, minBound=None, maxBound=None
):
"""
Load Images from OCP (Open Connectome Project).
The OCP is a web service for access to EM brain images and other neural image data.
The web-service can be accessed at http://www.openconnectomeproject.org/.
Parameters
----------
bucketName: string
Token name for the project in OCP. This name should exist on the server from which data is loaded.
resolution: nonnegative int
Resolution of the data in OCP
server: string, optional, default = 'ocp.me'
Name of the OCP server with the specified token.
startIdx: nonnegative int, optional, default = None
Convenience parameters to read only a subset of input files. Uses python slice conventions
(zero-based indexing with exclusive final position).
stopIdx: nonnegative int, optional
See startIdx.
minBound, maxBound: tuple of nonnegative int, optional, default = None
X,Y,Z bounds of the data to fetch from OCP. minBound contains the (xMin,yMin,zMin) while
maxBound contains (xMax,yMax,zMax).
Returns
-------
data: thunder.rdds.Images
An Images object, wrapping an RDD of with (int) : (numpy array) pairs
"""
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
# Checking StartIdx is smaller or equal to StopIdx
if startIdx is not None and stopIdx is not None and startIdx > stopIdx:
raise Exception("Error. startIdx {} is larger than stopIdx {}".format(startIdx, stopIdx))
data = loader.fromOCP(
bucketName,
resolution=resolution,
server=server,
startIdx=startIdx,
stopIdx=stopIdx,
minBound=minBound,
maxBound=maxBound,
)
return data
def test_crop(self):
dims = (2, 2, 4)
sz = reduce(lambda x, y: x*y, dims)
origAry = arange(sz, dtype='int16').reshape(dims)
imageData = ImagesLoader(self.sc).fromArrays([origAry])
croppedData = imageData.crop((0, 0, 0), (2, 2, 2))
crop = croppedData.collect()[0][1]
expected = squeeze(origAry[slice(0, 2), slice(0, 2), slice(0, 2)])
assert_true(array_equal(expected, crop))
assert_equals(tuple(expected.shape), croppedData._dims.count)
assert_equals(str(expected.dtype), croppedData._dtype)
def test_meanWithSingleRegionIndices3D(self):
ary1 = array([[[3, 5, 3], [6, 8, 6]], [[3, 5, 3], [6, 8, 6]]], dtype="int32")
ary2 = array([[[13, 15, 13], [16, 18, 16]], [[13, 15, 13], [16, 18, 16]]], dtype="int32")
images = ImagesLoader(self.sc).fromArrays([ary1, ary2])
indices = [[(1, 1, 1), (0, 0, 0)]] # one region with two indices
regionMeanImages = images.meanByRegions(indices)
self.__checkAttrPropagation(regionMeanImages, (1, 1))
collected = regionMeanImages.collect()
# check values
assert_equals(5, collected[0][1][0])
assert_equals(15, collected[1][1][0])
def test_subtract(self):
narys = 3
arys, sh, sz = _generateTestArrays(narys)
subVals = [1, arange(sz, dtype=dtypeFunc("int16")).reshape(sh)]
imageData = ImagesLoader(self.sc).fromArrays(arys)
for subVal in subVals:
subData = imageData.subtract(subVal)
subtracted = subData.collect()
expectedArys = map(lambda ary: ary - subVal, arys)
for actual, expected in zip(subtracted, expectedArys):
assert_true(allclose(expected, actual[1]))
def test_toBlocksWithSplit(self):
ary = arange(8, dtype=dtypeFunc("int16")).reshape((2, 4))
image = ImagesLoader(self.sc).fromArrays(ary)
groupedblocks = image.toBlocks((1, 2), units="s")
# collectedblocks = blocks.collect()
collectedgroupedblocks = groupedblocks.collect()
assert_equals((0, 0), collectedgroupedblocks[0][0].spatialKey)
assert_true(array_equal(ary[:, :2].ravel(), collectedgroupedblocks[0][1].ravel()))
assert_equals((0, 2), collectedgroupedblocks[1][0].spatialKey)
assert_true(array_equal(ary[:, 2:].ravel(), collectedgroupedblocks[1][1].ravel()))
def test_stats(self):
from test_utils import elementwiseMean, elementwiseVar
arys, shape, size = _generateTestArrays(2, 'uint8')
imageData = ImagesLoader(self.sc).fromArrays(arys)
statsval = imageData.stats()
floatarys = [ary.astype('float16') for ary in arys]
# StatsCounter contains a few different measures, only test a couple:
expectedMean = elementwiseMean(floatarys)
expectedVar = elementwiseVar(floatarys)
assert_true(allclose(expectedMean, statsval.mean()))
assert_true(allclose(expectedVar, statsval.variance()))
def test_planes(self):
dims = (2, 2, 4)
sz = reduce(lambda x, y: x * y, dims)
origAry = arange(sz, dtype=dtypeFunc("int16")).reshape(dims)
imageData = ImagesLoader(self.sc).fromArrays([origAry])
planedData = imageData.planes(0, 2)
planed = planedData.collect()[0][1]
expected = squeeze(origAry[slice(None), slice(None), slice(0, 2)])
assert_true(array_equal(expected, planed))
assert_equals(tuple(expected.shape), planedData._dims.count)
assert_equals(str(expected.dtype), planedData._dtype)
def test_crosscorrVolume(self):
random.seed(42)
ref = random.randn(25, 25, 3)
im = shift(ref, [2, -2, 0], mode='constant', order=0)
imIn = ImagesLoader(self.sc).fromArrays(im)
# use 3D cross correlation
paramOut = Registration('crosscorr').prepare(ref).fit(imIn).transformations[0].delta
imOut = Registration('crosscorr').prepare(ref).run(imIn).first()[1]
assert_true(allclose(paramOut, [2, -2, 0]))
assert_true(allclose(ref[:-2, 2:, :], imOut[:-2, 2:, :]))
# use 2D cross correlation on each plane
paramOut = Registration('planarcrosscorr').prepare(ref).fit(imIn).transformations[0].delta
imOut = Registration('planarcrosscorr').prepare(ref).run(imIn).first()[1]
assert_true(allclose(paramOut, [[2, -2], [2, -2], [2, -2]]))
assert_true(allclose(ref[:-2, 2:, :], imOut[:-2, 2:, :]))
def test_saveAsBinarySeries(self):
narys = 3
arys, aryShape, _ = _generateTestArrays(narys)
outdir = os.path.join(self.outputdir, "anotherdir")
os.mkdir(outdir)
assert_raises(
ValueError,
ImagesLoader(self.sc).fromArrays(arys).toBlocks(
(1, 1, 1), units="s").saveAsBinarySeries, outdir)
groupingDims = xrange(len(aryShape))
dtypes = ('int16', 'int32', 'float32')
paramIters = itertools.product(groupingDims, dtypes)
for idx, params in enumerate(paramIters):
gd, dt = params
self._run_tstSaveAsBinarySeries(idx, narys, dt, gd)
def _run_tst_filter(self, dataFunc, filterFunc):
narys = 3
arys, sh, sz = _generateTestArrays(narys)
sigma = 2
imageData = ImagesLoader(self.sc).fromArrays(arys)
filteredData = dataFunc(imageData, sigma)
filtered = filteredData.collect()
expectedArys = map(
lambda ary: TestImagesMethods._run_filter(ary, filterFunc, sigma),
arys)
for actual, expected in zip(filtered, expectedArys):
assert_true(allclose(expected, actual[1]))
assert_equals(tuple(expectedArys[0].shape), filtered[0][1].shape)
assert_equals(tuple(expectedArys[0].shape), filteredData._dims.count)
assert_equals(str(arys[0].dtype), str(filtered[0][1].dtype))
assert_equals(str(filtered[0][1].dtype), filteredData._dtype)
def test_reference_2d(self):
random.seed(42)
im0 = random.random_integers(0, high=127,
size=(25, 25)).astype('uint16')
im1 = random.random_integers(0, high=127,
size=(25, 25)).astype('uint16')
im2 = random.random_integers(0, high=127,
size=(25, 25)).astype('uint16')
imgIn = ImagesLoader(self.sc).fromArrays([im0, im1, im2])
ref = Register.reference(imgIn)
assert (allclose(ref, (im0 + im1 + im2) / 3))
ref = Register.reference(imgIn, startIdx=0, stopIdx=2)
assert (allclose(ref, (im0 + im1) / 2))
ref = Register.reference(imgIn, startIdx=1, stopIdx=2)
assert (allclose(ref, im1))
def test_reference_2d(self):
# test default reference calculation in 2D
random.seed(42)
im0 = random.rand(25, 25).astype('float')
im1 = random.rand(25, 25).astype('float')
im2 = random.rand(25, 25).astype('float')
imin = ImagesLoader(self.sc).fromArrays([im0, im1, im2])
reg = Registration('crosscorr').prepare(imin)
assert (allclose(reg.reference, (im0 + im1 + im2) / 3))
reg = Registration('crosscorr').prepare(imin, startidx=0, stopidx=2)
assert (allclose(reg.reference, (im0 + im1) / 2))
reg = Registration('crosscorr').prepare(imin, startidx=1, stopidx=2)
assert (allclose(reg.reference, im1))
def _run_tst_multitif(self, filename, expectedDtype):
imagePath = os.path.join(self.testResourcesDir, "multilayer_tif", filename)
tiffImages = ImagesLoader(self.sc).fromTif(imagePath).collect()
expectedNum = 1
expectedShape = (70, 75, 3) # 3 concatenated pages, each with single luminance channel
# 3 images have increasing #s of black dots, so lower luminance overall
expectedSums = [1140006, 1119161, 1098917]
expectedKey = 0
assert_equals(expectedNum, len(tiffImages), "Expected %d images, got %d" % (expectedNum, len(tiffImages)))
tiffImage = tiffImages[0]
assert_equals(expectedKey, tiffImage[0], "Expected key %s, got %s" % (str(expectedKey), str(tiffImage[0])))
assert_true(isinstance(tiffImage[1], ndarray),
"Value type error; expected image value to be numpy ndarray, was " + str(type(tiffImage[1])))
assert_equals(expectedDtype, str(tiffImage[1].dtype))
assert_equals(expectedShape, tiffImage[1].shape)
for channelidx in xrange(0, expectedShape[2]):
assert_equals(expectedSums[channelidx], tiffImage[1][:, :, channelidx].flatten().sum())
def test_fromMultipageTif(self):
imagepath = os.path.join(self.testresourcesdir, "multilayer_tif", "dotdotdot_lzw.tif")
tifimages = ImagesLoader(self.sc).fromMultipageTif(imagepath, self.sc).collect()
expectednum = 1
expectedshape = (70, 75, 3) # 3 concatenated pages, each with single luminance channel
# 3 images have increasing #s of black dots, so lower luminance overall
expectedsums = [1140006, 1119161, 1098917]
expectedkey = 0
#self._evaluateMultipleImages(tifimages, expectednum, expectedshape, expectedkeys, expectedsums)
assert_equals(expectednum, len(tifimages), "Expected %s images, got %d" % (expectednum, len(tifimages)))
tifimage = tifimages[0]
assert_equals(expectedkey, tifimage[0], "Expected key %s, got %s" % (str(expectedkey), str(tifimage[0])))
assert_true(isinstance(tifimage[1], ndarray),
"Value type error; expected image value to be numpy ndarray, was " + str(type(tifimage[1])))
assert_equals('uint8', str(tifimage[1].dtype))
assert_equals(expectedshape, tifimage[1].shape)
for channelidx in xrange(0, expectedshape[2]):
assert_equals(expectedsums[channelidx], tifimage[1][:, :, channelidx].flatten().sum())
def test_save_load(self):
# test basic saving a loading functionality
# new registration methods should add tests
# for loading and saving
random.seed(42)
ref = random.randn(25, 25)
im = shift(ref, [2, 0], mode='constant', order=0)
imin = ImagesLoader(self.sc).fromArrays(im)
reg = Registration('crosscorr')
reg.prepare(ref)
model1 = reg.fit(imin)
t = tempfile.mkdtemp()
model1.save(t + '/test.json')
model2 = Registration.load(t + '/test.json')
out1 = model1.transform(imin).first()[1]
out2 = model2.transform(imin).first()[1]
assert (allclose(out1, out2))
def loadImages(self, dataPath, dims=None, dtype=None, inputFormat='stack', ext=None,
startIdx=None, stopIdx=None, recursive=False, nplanes=None, npartitions=None,
renumber=False, confFilename='conf.json'):
"""
Loads an Images object from data stored as a binary image stack, tif, or png files.
Supports single files or multiple files, stored on a local file system, a networked file sytem
(mounted and available on all nodes), or Amazon S3. HDFS is not currently supported for image file data.
Parameters
----------
dataPath: string
Path to data files or directory, as either a local filesystem path or a URI.
May include a single '*' wildcard in the filename. Examples of valid dataPaths include
'local/directory/*.stack", "s3n:///my-s3-bucket/data/", or "file:///mnt/another/directory/".
dims: tuple of positive int, optional (required if inputFormat is 'stack')
Image dimensions. Binary stack data will be interpreted as a multidimensional array
with the given dimensions, and should be stored in row-major order (Fortran or Matlab convention),
where the first dimension changes most rapidly. For 'png' or 'tif' data dimensions
will be read from the image file headers.
inputFormat: str, optional, default = 'stack'
Expected format of the input data: 'stack', 'png', or 'tif'. 'stack' indicates flat binary stacks.
'png' or 'tif' indicate image format. Page of a multipage tif file will be extend along
the third dimension. Separate files interpreted as distinct records, with ordering
given by lexicographic sorting of file names.
ext: string, optional, default = None
File extension, default will be "bin" if inputFormat=="stack", "tif" for inputFormat=='tif',
and 'png' for inputFormat=="png".
dtype: string or numpy dtype, optional, default = 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string.
Ignored for 'tif' or 'png' (data will be inferred from image formats).
startIdx: nonnegative int, optional, default = None
Convenience parameters to read only a subset of input files. Uses python slice conventions
(zero-based indexing with exclusive final position). These parameters give the starting
and final index after lexicographic sorting.
stopIdx: nonnegative int, optional, default = None
See startIdx.
recursive: boolean, optional, default = False
If true, will recursively descend directories rooted at dataPath, loading all files
in the tree with an appropriate extension.
nplanes: positive integer, optional, default = None
Subdivide individual image files. Every `nplanes` from each file will be considered a new record.
With nplanes=None (the default), a single file will be considered as representing a single record.
If the number of records per file is not the same across all files, then `renumber` should be set
to True to ensure consistent keys.
npartitions: positive int, optional, default = None
Specify number of partitions for the RDD, if unspecified will use 1 partition per image.
renumber: boolean, optional, default = False
Recalculate keys for records after images are loading. Only necessary if different files contain
different number of records (e.g. due to specifying nplanes). See Images.renumber().
confFilename : string, optional, default = 'conf.json'
Name of conf file if using to specify parameters for binary stack data
Returns
-------
data: thunder.rdds.Images
An Images object, wrapping an RDD of with (int) : (numpy array) pairs
"""
checkParams(inputFormat, ['stack', 'png', 'tif', 'tif-stack'])
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
# Checking StartIdx is smaller or equal to StopIdx
if startIdx is not None and stopIdx is not None and startIdx > stopIdx:
raise Exception("Error. startIdx {} is larger than stopIdx {}".inputFormat(startIdx, stopIdx))
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if inputFormat.lower() == 'stack':
data = loader.fromStack(dataPath, dims=dims, dtype=dtype, ext=ext, startIdx=startIdx, stopIdx=stopIdx,
recursive=recursive, nplanes=nplanes, npartitions=npartitions,
confFilename=confFilename)
elif inputFormat.lower().startswith('tif'):
data = loader.fromTif(dataPath, ext=ext, startIdx=startIdx, stopIdx=stopIdx, recursive=recursive,
nplanes=nplanes, npartitions=npartitions)
else:
if nplanes:
raise NotImplementedError("nplanes argument is not supported for png files")
data = loader.fromPng(dataPath, ext=ext, startIdx=startIdx, stopIdx=stopIdx, recursive=recursive,
npartitions=npartitions)
if not renumber:
return data
else:
return data.renumber()
def _run_tstSaveAsBinarySeries(self, testIdx, narys_, valDtype,
groupingDim_):
"""Pseudo-parameterized test fixture, allows reusing existing spark context
"""
paramStr = "(groupingdim=%d, valuedtype='%s')" % (groupingDim_,
valDtype)
arys, aryShape, arySize = _generateTestArrays(narys_, dtype_=valDtype)
dims = aryShape[:]
outdir = os.path.join(self.outputdir, "anotherdir%02d" % testIdx)
images = ImagesLoader(self.sc).fromArrays(arys)
slicesPerDim = [1] * arys[0].ndim
slicesPerDim[groupingDim_] = arys[0].shape[groupingDim_]
images.toBlocks(slicesPerDim,
units="splits").saveAsBinarySeries(outdir)
ndims = len(aryShape)
# prevent padding to 4-byte boundaries: "=" specifies no alignment
unpacker = struct.Struct('=' + 'h' * ndims +
dtypeFunc(valDtype).char * narys_)
def calcExpectedNKeys():
tmpShape = list(dims[:])
del tmpShape[groupingDim_]
return prod(tmpShape)
expectedNKeys = calcExpectedNKeys()
def byrec(f_, unpacker_, nkeys_):
rec = True
while rec:
rec = f_.read(unpacker_.size)
if rec:
allRecVals = unpacker_.unpack(rec)
yield allRecVals[:nkeys_], allRecVals[nkeys_:]
outFilenames = glob.glob(os.path.join(outdir, "*.bin"))
assert_equals(dims[groupingDim_], len(outFilenames))
for outFilename in outFilenames:
with open(outFilename, 'rb') as f:
nkeys = 0
for keys, vals in byrec(f, unpacker, ndims):
nkeys += 1
assert_equals(narys_, len(vals))
for valIdx, val in enumerate(vals):
assert_equals(
arys[valIdx][keys], val,
"Expected %g, got %g, for test %d %s" %
(arys[valIdx][keys], val, testIdx, paramStr))
assert_equals(expectedNKeys, nkeys)
confName = os.path.join(outdir, "conf.json")
assert_true(os.path.isfile(confName))
with open(os.path.join(outdir, "conf.json"), 'r') as fconf:
import json
conf = json.load(fconf)
assert_equals(outdir, conf['input'])
assert_equals(len(aryShape), conf['nkeys'])
assert_equals(narys_, conf['nvalues'])
assert_equals(valDtype, conf['valuetype'])
assert_equals('int16', conf['keytype'])
assert_true(os.path.isfile(os.path.join(outdir, 'SUCCESS')))
def test_min(self):
from numpy import minimum
arys, shape, size = _generateTestArrays(2, 'uint8')
imageData = ImagesLoader(self.sc).fromArrays(arys)
minVal = imageData.min()
assert_true(array_equal(reduce(minimum, arys), minVal))
def convertImagesToSeries(self,
datapath,
outputdirpath,
dims=None,
inputformat='stack',
dtype='int16',
blocksize="150M",
startidx=None,
stopidx=None,
shuffle=False,
overwrite=False):
"""
Write out Images data as Series data, saved in a flat binary format.
The resulting Series data files may subsequently be read in using the loadSeries() method. The Series data
object that results will be equivalent to that which would be generated by loadImagesAsSeries(). It is expected
that loading Series data directly from the series flat binary format, using loadSeries(), will be faster than
converting image data to a Series object through loadImagesAsSeries().
Parameters
----------
datapath: string
Path to data files or directory, specified as either a local filesystem path or in a URI-like format,
including scheme. A datapath argument may include a single '*' wildcard character in the filename. Examples
of valid datapaths include 'a/local/relative/directory/*.stack", "s3n:///my-s3-bucket/data/mydatafile.tif",
"/mnt/my/absolute/data/directory/", or "file:///mnt/another/data/directory/".
outputdirpath: string
Path to a directory into which to write Series file output. An outputdir argument may be either a path
on the local file system or a URI-like format, as in datapath. Examples of valid outputdirpaths include
"a/relative/directory/", "s3n:///my-s3-bucket/data/myoutput/", or "file:///mnt/a/new/directory/". If the
directory specified by outputdirpath already exists and the 'overwrite' parameter is False, this method
will throw a ValueError. If the directory exists and 'overwrite' is True, the existing directory and all
its contents will be deleted and overwritten.
dims: tuple of positive int, optional (but required if inputformat is 'stack')
Dimensions of input image data, for instance (1024, 1024, 48). Binary stack data will be interpreted as
coming from a multidimensional array of the specified dimensions.
The first dimension of the passed dims tuple should be the one that is changing most rapidly
on disk. So for instance given dims of (x, y, z), the coordinates of the data in a binary stack file
should be ordered as [(x0, y0, z0), (x1, y0, z0), ..., (xN, y0, z0), (x0, y1, z0), (x1, y1, z0), ...,
(xN, yM, z0), (x0, y0, z1), ..., (xN, yM, zP)]. This is the opposite convention from that used by numpy,
which by default has the fastest-changing dimension listed last (column-major convention). Thus, if loading
a numpy array `ary`, where `ary.shape == (z, y, x)`, written to disk by `ary.tofile("myarray.stack")`, the
corresponding dims parameter should be (x, y, z).
If inputformat is 'tif-stack', the dims parameter (if any) will be ignored; data dimensions will instead
be read out from the tif file headers.
inputformat: {'stack', 'tif-stack'}. optional, default 'stack'
Expected format of the input data. 'stack' indicates flat files of raw binary data, while 'tif-stack'
indicates a sequence of multipage tif files, with each page of the tif corresponding to a separate z-plane.
For both stacks and tif stacks, separate files are interpreted as distinct time points, with ordering
given by lexicographic sorting of file names.
This method assumes that stack data consists of signed 16-bit integers in native byte order. The lower-level
API method SeriesLoader.saveFromStack() allows alternative data types to be read in.
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string. If inputformat is
'tif-stack', the dtype parameter (if any) will be ignored; data type will instead be read out from the
tif headers.
blocksize: string formatted as e.g. "64M", "512k", "2G", or positive int. optional, default "150M"
Requested size of individual output files in bytes (or kilobytes, megabytes, gigabytes). This parameter
also indirectly controls the number of Spark partitions to be used, with one partition used per block
created.
startidx: nonnegative int, optional
startidx and stopidx are convenience parameters to allow only a subset of input files to be read in. These
parameters give the starting index (inclusive) and final index (exclusive) of the data files to be used
after lexicographically sorting all input data files matching the datapath argument. For example,
startidx=None (the default) and stopidx=10 will cause only the first 10 data files in datapath to be read
in; startidx=2 and stopidx=3 will cause only the third file (zero-based index of 2) to be read in. startidx
and stopidx use the python slice indexing convention (zero-based indexing with an exclusive final position).
stopidx: nonnegative int, optional
See startidx.
shuffle: boolean, optional, default False
Controls whether the conversion from Images to Series formats will make use of a Spark shuffle-based method.
The default at present is not to use a shuffle. The shuffle-based method may lead to higher performance in
some cases, but the default method appears to be more stable with larger data set sizes. This default may
change in future releases.
overwrite: boolean, optional, default False
If true, the directory specified by outputdirpath will first be deleted, along with all its contents, if it
already exists. (Use with caution.) If false, a ValueError will be thrown if outputdirpath is found to
already exist.
"""
checkparams(inputformat, ['stack', 'tif-stack'])
if inputformat.lower() == 'stack' and not dims:
raise ValueError(
"Dimensions ('dims' parameter) must be specified if loading from binary image stack"
+ " ('stack' value for 'inputformat' parameter)")
if shuffle:
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputformat.lower() == 'stack':
loader.fromStack(datapath, dims, dtype=dtype, startidx=startidx, stopidx=stopidx)\
.saveAsBinarySeries(outputdirpath, blockSize=blocksize, overwrite=overwrite)
else:
loader.fromMultipageTif(datapath, startidx=startidx, stopidx=stopidx)\
.saveAsBinarySeries(outputdirpath, blockSize=blocksize, overwrite=overwrite)
else:
from thunder.rdds.fileio.seriesloader import SeriesLoader
loader = SeriesLoader(self._sc)
if inputformat.lower() == 'stack':
loader.saveFromStack(datapath,
outputdirpath,
dims,
datatype=dtype,
blockSize=blocksize,
overwrite=overwrite,
startidx=startidx,
stopidx=stopidx)
else:
loader.saveFromMultipageTif(datapath,
outputdirpath,
blockSize=blocksize,
startidx=startidx,
stopidx=stopidx,
overwrite=overwrite)
def test_max(self):
from numpy import maximum
arys, shape, size = _generate_test_arrays(2, 'uint8')
imagedata = ImagesLoader(self.sc).fromArrays(arys)
maxval = imagedata.max()
assert_true(array_equal(reduce(maximum, arys), maxval))
def loadImagesAsSeries(self,
dataPath,
dims=None,
inputFormat='stack',
ext=None,
dtype='int16',
blockSize="150M",
blockSizeUnits="pixels",
startIdx=None,
stopIdx=None,
shuffle=True,
recursive=False):
"""
Load Images data as Series data.
Parameters
----------
dataPath: string
Path to data files or directory, specified as either a local filesystem path or in a URI-like format,
including scheme. A dataPath argument may include a single '*' wildcard character in the filename. Examples
of valid dataPaths include 'a/local/relative/directory/*.stack", "s3n:///my-s3-bucket/data/mydatafile.tif",
"/mnt/my/absolute/data/directory/", or "file:///mnt/another/data/directory/".
dims: tuple of positive int, optional (but required if inputFormat is 'stack')
Dimensions of input image data, for instance (1024, 1024, 48). Binary stack data will be interpreted as
coming from a multidimensional array of the specified dimensions.
The first dimension of the passed dims tuple should be the one that is changing most rapidly
on disk. So for instance given dims of (x, y, z), the coordinates of the data in a binary stack file
should be ordered as [(x0, y0, z0), (x1, y0, z0), ..., (xN, y0, z0), (x0, y1, z0), (x1, y1, z0), ...,
(xN, yM, z0), (x0, y0, z1), ..., (xN, yM, zP)]. This is the opposite convention from that used by numpy,
which by default has the fastest-changing dimension listed last (column-major convention). Thus, if loading
a numpy array `ary`, where `ary.shape == (z, y, x)`, written to disk by `ary.tofile("myarray.stack")`, the
corresponding dims parameter should be (x, y, z).
If inputFormat is 'tif', the dims parameter (if any) will be ignored; data dimensions will instead
be read out from the tif file headers.
inputFormat: {'stack', 'tif'}. optional, default 'stack'
Expected format of the input data. 'stack' indicates flat files of raw binary data, while 'tif' indicates
greyscale / luminance TIF images. Each page of a multipage tif file will be interpreted as a separate
z-plane. For both stacks and tif stacks, separate files are interpreted as distinct time points, with
ordering given by lexicographic sorting of file names.
ext: string, optional, default None
Extension required on data files to be loaded. By default will be "stack" if inputFormat=="stack", "tif" for
inputFormat=='tif'.
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string. If inputFormat is
'tif', the dtype parameter (if any) will be ignored; data type will instead be read out from the
tif headers.
blockSize: string formatted as e.g. "64M", "512k", "2G", or positive int. optional, default "150M"
Requested size of individual output files in bytes (or kilobytes, megabytes, gigabytes). If shuffle=True,
blockSize can also be a tuple of int specifying either the number of pixels or of splits per dimension to
apply to the loaded images, or an instance of BlockingStrategy. Whether a tuple of int is interpreted as
pixels or as splits depends on the value of the blockSizeUnits parameter. blockSize also indirectly
controls the number of Spark partitions to be used, with one partition used per block created.
blockSizeUnits: string, either "pixels" or "splits" (or unique prefix of each, such as "s"), default "pixels"
Specifies units to be used in interpreting a tuple passed as blockSizeSpec when shuffle=True. If a string
or a BlockingStrategy instance is passed as blockSizeSpec, or if shuffle=False, this parameter has no
effect.
startIdx: nonnegative int, optional
startIdx and stopIdx are convenience parameters to allow only a subset of input files to be read in. These
parameters give the starting index (inclusive) and final index (exclusive) of the data files to be used
after lexicographically sorting all input data files matching the dataPath argument. For example,
startIdx=None (the default) and stopIdx=10 will cause only the first 10 data files in dataPath to be read
in; startIdx=2 and stopIdx=3 will cause only the third file (zero-based index of 2) to be read in. startIdx
and stopIdx use the python slice indexing convention (zero-based indexing with an exclusive final position).
stopIdx: nonnegative int, optional
See startIdx.
shuffle: boolean, optional, default True
Controls whether the conversion from Images to Series formats will make use of a Spark shuffle-based method.
recursive: boolean, default False
If true, will recursively descend directories rooted at dataPath, loading all files in the tree that
have an appropriate extension. Recursive loading is currently only implemented for local filesystems
(not s3), and only with shuffle=True.
Returns
-------
data: thunder.rdds.Series
A newly-created Series object, wrapping an RDD of timeseries data generated from the images in dataPath.
This RDD will have as keys an n-tuple of int, with n given by the dimensionality of the original images. The
keys will be the zero-based spatial index of the timeseries data in the RDD value. The value will be
a numpy array of length equal to the number of image files loaded. Each loaded image file will contribute
one point to this value array, with ordering as implied by the lexicographic ordering of image file names.
"""
checkParams(inputFormat, ['stack', 'tif', 'tif-stack'])
if inputFormat.lower() == 'stack' and not dims:
raise ValueError(
"Dimensions ('dims' parameter) must be specified if loading from binary image stack"
+ " ('stack' value for 'inputFormat' parameter)")
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if shuffle:
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputFormat.lower() == 'stack':
images = loader.fromStack(dataPath,
dims,
dtype=dtype,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive)
else:
# tif / tif stack
images = loader.fromTif(dataPath,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive)
return images.toBlocks(blockSize, units=blockSizeUnits).toSeries()
else:
from thunder.rdds.fileio.seriesloader import SeriesLoader
loader = SeriesLoader(self._sc)
if inputFormat.lower() == 'stack':
return loader.fromStack(dataPath,
dims,
ext=ext,
dtype=dtype,
blockSize=blockSize,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive)
else:
# tif / tif stack
return loader.fromTif(dataPath,
ext=ext,
blockSize=blockSize,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive)
def setUp(self):
super(TestImagesMeanByRegions, self).setUp()
self.ary1 = array([[3, 5], [6, 8]], dtype='int32')
self.ary2 = array([[13, 15], [16, 18]], dtype='int32')
self.images = ImagesLoader(self.sc).fromArrays([self.ary1, self.ary2])
def loadImages(self,
datapath,
dims=None,
inputformat='stack',
dtype='int16',
startidx=None,
stopidx=None):
"""
Loads an Images object from data stored as a binary image stack, tif, tif-stack, or png files.
Supports single files or multiple files, stored on a local file system, a networked file sytem
(mounted and available on all nodes), or Amazon S3. HDFS is not currently supported for image file data.
Parameters
----------
datapath: string
Path to data files or directory, specified as either a local filesystem path or in a URI-like format,
including scheme. A datapath argument may include a single '*' wildcard character in the filename. Examples
of valid datapaths include 'a/local/relative/directory/*.stack", "s3n:///my-s3-bucket/data/mydatafile.tif",
"/mnt/my/absolute/data/directory/", or "file:///mnt/another/data/directory/".
dims: tuple of positive int, optional (but required if inputformat is 'stack')
Dimensions of input image data, similar to a numpy 'shape' parameter, for instance (1024, 1024, 48). Binary
stack data will be interpreted as coming from a multidimensional array of the specified dimensions. Stack
data should be stored in row-major order (Fortran or Matlab convention) rather than column-major order (C
or python/numpy convention), where the first dimension corresponds to that which is changing most rapidly
on disk. So for instance given dims of (x, y, z), the coordinates of the data in a binary stack file
should be ordered as [(x0, y0, z0), (x1, y0, zo), ..., (xN, y0, z0), (x0, y1, z0), (x1, y1, z0), ...,
(xN, yM, z0), (x0, y0, z1), ..., (xN, yM, zP)].
If inputformat is 'png', 'tif', or'tif-stack', the dims parameter (if any) will be ignored; data dimensions
will instead be read out from the image file headers.
inputformat: {'stack', 'png', 'tif', 'tif-stack'}. optional, default 'stack'
Expected format of the input data. 'stack' indicates flat files of raw binary data. 'png' or 'tif' indicate
two-dimensional image files of the corresponding formats. 'tif-stack' indicates a sequence of multipage tif
files, with each page of the tif corresponding to a separate z-plane.
For all formats, separate files are interpreted as distinct time points, with ordering given by
lexicographic sorting of file names.
This method assumes that stack data consists of signed 16-bit integers in native byte order. Data types of
image file data will be as specified in the file headers.
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string. If inputformat is
'tif-stack', the dtype parameter (if any) will be ignored; data type will instead be read out from the
tif headers.
startidx: nonnegative int, optional
startidx and stopidx are convenience parameters to allow only a subset of input files to be read in. These
parameters give the starting index (inclusive) and final index (exclusive) of the data files to be used
after lexicographically sorting all input data files matching the datapath argument. For example,
startidx=None (the default) and stopidx=10 will cause only the first 10 data files in datapath to be read
in; startidx=2 and stopidx=3 will cause only the third file (zero-based index of 2) to be read in. startidx
and stopidx use the python slice indexing convention (zero-based indexing with an exclusive final position).
stopidx: nonnegative int, optional
See startidx.
Returns
-------
data: thunder.rdds.Images
A newly-created Images object, wrapping an RDD of <int index, numpy array> key-value pairs.
"""
checkparams(inputformat, ['stack', 'png', 'tif', 'tif-stack'])
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputformat.lower() == 'stack':
data = loader.fromStack(datapath,
dims,
dtype=dtype,
startidx=startidx,
stopidx=stopidx)
elif inputformat.lower() == 'tif':
data = loader.fromTif(datapath, startidx=startidx, stopidx=stopidx)
elif inputformat.lower() == 'tif-stack':
data = loader.fromMultipageTif(datapath,
startidx=startidx,
stopidx=stopidx)
else:
data = loader.fromPng(datapath)
return data
def loadImages(self,
dataPath,
dims=None,
dtype=None,
inputFormat='stack',
ext=None,
startIdx=None,
stopIdx=None,
recursive=False,
nplanes=None,
npartitions=None,
renumber=False,
confFilename='conf.json'):
"""
Loads an Images object from data stored as a binary image stack, tif, or png files.
Supports single files or multiple files, stored on a local file system, a networked file sytem
(mounted and available on all nodes), or Amazon S3. HDFS is not currently supported for image file data.
Parameters
----------
dataPath: string
Path to data files or directory, specified as either a local filesystem path or in a URI-like format,
including scheme. A dataPath argument may include a single '*' wildcard character in the filename. Examples
of valid dataPaths include 'a/local/relative/directory/*.stack", "s3n:///my-s3-bucket/data/mydatafile.tif",
"/mnt/my/absolute/data/directory/", or "file:///mnt/another/data/directory/".
dims: tuple of positive int, optional (but required if inputFormat is 'stack')
Dimensions of input image data, similar to a numpy 'shape' parameter, for instance (1024, 1024, 48). Binary
stack data will be interpreted as coming from a multidimensional array of the specified dimensions. Stack
data should be stored in row-major order (Fortran or Matlab convention) rather than column-major order (C
or python/numpy convention), where the first dimension corresponds to that which is changing most rapidly
on disk. So for instance given dims of (x, y, z), the coordinates of the data in a binary stack file
should be ordered as [(x0, y0, z0), (x1, y0, zo), ..., (xN, y0, z0), (x0, y1, z0), (x1, y1, z0), ...,
(xN, yM, z0), (x0, y0, z1), ..., (xN, yM, zP)].
If inputFormat is 'png' or 'tif', the dims parameter (if any) will be ignored; data dimensions
will instead be read out from the image file headers.
inputFormat: {'stack', 'png', 'tif'}. optional, default 'stack'
Expected format of the input data. 'stack' indicates flat files of raw binary data. 'png' or 'tif' indicate
image files of the corresponding formats. Each page of a multipage tif file will be interpreted as a
separate z-plane. For all formats, separate files are interpreted as distinct time points, with ordering
given by lexicographic sorting of file names.
ext: string, optional, default None
Extension required on data files to be loaded. By default will be "stack" if inputFormat=="stack", "tif" for
inputFormat=='tif', and 'png' for inputFormat="png".
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string. If inputFormat is
'tif' or 'png', the dtype parameter (if any) will be ignored; data type will instead be read out from the
tif headers.
startIdx: nonnegative int, optional
startIdx and stopIdx are convenience parameters to allow only a subset of input files to be read in. These
parameters give the starting index (inclusive) and final index (exclusive) of the data files to be used
after lexicographically sorting all input data files matching the dataPath argument. For example,
startIdx=None (the default) and stopIdx=10 will cause only the first 10 data files in dataPath to be read
in; startIdx=2 and stopIdx=3 will cause only the third file (zero-based index of 2) to be read in. startIdx
and stopIdx use the python slice indexing convention (zero-based indexing with an exclusive final position).
stopIdx: nonnegative int, optional
See startIdx.
recursive: boolean, default False
If true, will recursively descend directories rooted at dataPath, loading all files in the tree that
have an appropriate extension. Recursive loading is currently only implemented for local filesystems
(not s3).
nplanes: positive integer, default None
If passed, will cause a single image file to be subdivided into multiple records. Every
`nplanes` z-planes (or multipage tif pages) in the file will be taken as a new record, with the
first nplane planes of the first file being record 0, the second nplane planes being record 1, etc,
until the first file is exhausted and record ordering continues with the first nplane planes of the
second file, and so on. With nplanes=None (the default), a single file will be considered as
representing a single record. Keys are calculated assuming that all input files contain the same
number of records; if the number of records per file is not the same across all files,
then `renumber` should be set to True to ensure consistent keys.
npartitions: positive int, optional
If specified, request a certain number of partitions for the underlying Spark RDD. Default is 1
partition per image file.
renumber: boolean, optional, default False
If renumber evaluates to True, then the keys for each record will be explicitly recalculated after
all images are loaded. This should only be necessary at load time when different files contain
different number of records. See Images.renumber().
Returns
-------
data: thunder.rdds.Images
A newly-created Images object, wrapping an RDD of <int index, numpy array> key-value pairs.
"""
checkParams(inputFormat, ['stack', 'png', 'tif', 'tif-stack'])
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if inputFormat.lower() == 'stack':
data = loader.fromStack(dataPath,
dims=dims,
dtype=dtype,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive,
nplanes=nplanes,
npartitions=npartitions,
confFilename=confFilename)
elif inputFormat.lower().startswith('tif'):
data = loader.fromTif(dataPath,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive,
nplanes=nplanes,
npartitions=npartitions)
else:
if nplanes:
raise NotImplementedError(
"nplanes argument is not supported for png files")
data = loader.fromPng(dataPath,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive,
npartitions=npartitions)
if not renumber:
return data
else:
return data.renumber()
def convertImagesToSeries(self,
dataPath,
outputDirPath,
dims=None,
inputFormat='stack',
ext=None,
dtype='int16',
blockSize="150M",
blockSizeUnits="pixels",
startIdx=None,
stopIdx=None,
shuffle=True,
overwrite=False,
recursive=False,
nplanes=None,
npartitions=None,
renumber=False):
"""
Write out Images data as Series data, saved in a flat binary format.
The resulting Series data files may subsequently be read in using the loadSeries() method. The Series data
object that results will be equivalent to that which would be generated by loadImagesAsSeries(). It is expected
that loading Series data directly from the series flat binary format, using loadSeries(), will be faster than
converting image data to a Series object through loadImagesAsSeries().
Parameters
----------
dataPath: string
Path to data files or directory, specified as either a local filesystem path or in a URI-like format,
including scheme. A dataPath argument may include a single '*' wildcard character in the filename. Examples
of valid dataPaths include 'a/local/relative/directory/*.stack", "s3n:///my-s3-bucket/data/mydatafile.tif",
"/mnt/my/absolute/data/directory/", or "file:///mnt/another/data/directory/".
outputDirPath: string
Path to a directory into which to write Series file output. An outputdir argument may be either a path
on the local file system or a URI-like format, as in dataPath. Examples of valid outputDirPaths include
"a/relative/directory/", "s3n:///my-s3-bucket/data/myoutput/", or "file:///mnt/a/new/directory/". If the
directory specified by outputDirPath already exists and the 'overwrite' parameter is False, this method
will throw a ValueError. If the directory exists and 'overwrite' is True, the existing directory and all
its contents will be deleted and overwritten.
dims: tuple of positive int, optional (but required if inputFormat is 'stack')
Dimensions of input image data, for instance (1024, 1024, 48). Binary stack data will be interpreted as
coming from a multidimensional array of the specified dimensions.
The first dimension of the passed dims tuple should be the one that is changing most rapidly
on disk. So for instance given dims of (x, y, z), the coordinates of the data in a binary stack file
should be ordered as [(x0, y0, z0), (x1, y0, z0), ..., (xN, y0, z0), (x0, y1, z0), (x1, y1, z0), ...,
(xN, yM, z0), (x0, y0, z1), ..., (xN, yM, zP)]. This is the opposite convention from that used by numpy,
which by default has the fastest-changing dimension listed last (column-major convention). Thus, if loading
a numpy array `ary`, where `ary.shape == (z, y, x)`, written to disk by `ary.tofile("myarray.stack")`, the
corresponding dims parameter should be (x, y, z).
If inputFormat is 'tif', the dims parameter (if any) will be ignored; data dimensions will instead
be read out from the tif file headers.
inputFormat: {'stack', 'tif'}. optional, default 'stack'
Expected format of the input data. 'stack' indicates flat files of raw binary data, while 'tif' indicates
greyscale / luminance TIF images. Each page of a multipage tif file will be interpreted as a separate
z-plane. For both stacks and tif stacks, separate files are interpreted as distinct time points, with
ordering given by lexicographic sorting of file names.
ext: string, optional, default None
Extension required on data files to be loaded. By default will be "stack" if inputFormat=="stack", "tif" for
inputFormat=='tif'.
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string. If inputFormat is
'tif', the dtype parameter (if any) will be ignored; data type will instead be read out from the
tif headers.
blockSize: string formatted as e.g. "64M", "512k", "2G", or positive int, tuple of positive int, or instance of
BlockingStrategy. optional, default "150M"
Requested size of individual output files in bytes (or kilobytes, megabytes, gigabytes). blockSize can also
be an instance of blockingStrategy, or a tuple of int specifying either the number of pixels or of splits
per dimension to apply to the loaded images. Whether a tuple of int is interpreted as pixels or as splits
depends on the value of the blockSizeUnits parameter. This parameter also indirectly controls the number
of Spark partitions to be used, with one partition used per block created.
blockSizeUnits: string, either "pixels" or "splits" (or unique prefix of each, such as "s"), default "pixels"
Specifies units to be used in interpreting a tuple passed as blockSizeSpec when shuffle=True. If a string
or a BlockingStrategy instance is passed as blockSizeSpec, or if shuffle=False, this parameter has no
effect.
startIdx: nonnegative int, optional
startIdx and stopIdx are convenience parameters to allow only a subset of input files to be read in. These
parameters give the starting index (inclusive) and final index (exclusive) of the data files to be used
after lexicographically sorting all input data files matching the dataPath argument. For example,
startIdx=None (the default) and stopIdx=10 will cause only the first 10 data files in dataPath to be read
in; startIdx=2 and stopIdx=3 will cause only the third file (zero-based index of 2) to be read in. startIdx
and stopIdx use the python slice indexing convention (zero-based indexing with an exclusive final position).
stopIdx: nonnegative int, optional
See startIdx.
shuffle: boolean, optional, default True
Controls whether the conversion from Images to Series formats will make use of a Spark shuffle-based method.
overwrite: boolean, optional, default False
If true, the directory specified by outputDirPath will first be deleted, along with all its contents, if it
already exists. (Use with caution.) If false, a ValueError will be thrown if outputDirPath is found to
already exist.
recursive: boolean, default False
If true, will recursively descend directories rooted at dataPath, loading all files in the tree that
have an appropriate extension. Recursive loading is currently only implemented for local filesystems
(not s3), and only with shuffle=True.
nplanes: positive integer, default None
If passed, will cause a single image file to be subdivided into multiple records. Every
`nplanes` z-planes (or multipage tif pages) in the file will be taken as a new record, with the
first nplane planes of the first file being record 0, the second nplane planes being record 1, etc,
until the first file is exhausted and record ordering continues with the first nplane planes of the
second file, and so on. With nplanes=None (the default), a single file will be considered as
representing a single record. Keys are calculated assuming that all input files contain the same
number of records; if the number of records per file is not the same across all files,
then `renumber` should be set to True to ensure consistent keys. nplanes is only supported for
shuffle=True (the default).
npartitions: positive int, optional
If specified, request a certain number of partitions for the underlying Spark RDD. Default is 1
partition per image file. Only applies when shuffle=True.
renumber: boolean, optional, default False
If renumber evaluates to True, then the keys for each record will be explicitly recalculated after
all images are loaded. This should only be necessary at load time when different files contain
different number of records. renumber is only supported for shuffle=True (the default). See
Images.renumber().
"""
checkParams(inputFormat, ['stack', 'tif', 'tif-stack'])
if inputFormat.lower() == 'stack' and not dims:
raise ValueError(
"Dimensions ('dims' parameter) must be specified if loading from binary image stack"
+ " ('stack' value for 'inputFormat' parameter)")
if not overwrite:
raiseErrorIfPathExists(outputDirPath,
awsCredentialsOverride=self._credentials)
overwrite = True # prevent additional downstream checks for this path
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if shuffle:
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputFormat.lower() == 'stack':
images = loader.fromStack(dataPath,
dims,
ext=ext,
dtype=dtype,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive,
nplanes=nplanes,
npartitions=npartitions)
else:
# 'tif' or 'tif-stack'
images = loader.fromTif(dataPath,
ext=ext,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive,
nplanes=nplanes,
npartitions=npartitions)
if renumber:
images = images.renumber()
images.toBlocks(blockSize,
units=blockSizeUnits).saveAsBinarySeries(
outputDirPath, overwrite=overwrite)
else:
from thunder.rdds.fileio.seriesloader import SeriesLoader
if nplanes is not None:
raise NotImplementedError(
"nplanes is not supported with shuffle=False")
if npartitions is not None:
raise NotImplementedError(
"npartitions is not supported with shuffle=False")
loader = SeriesLoader(self._sc)
if inputFormat.lower() == 'stack':
loader.saveFromStack(dataPath,
outputDirPath,
dims,
ext=ext,
dtype=dtype,
blockSize=blockSize,
overwrite=overwrite,
startIdx=startIdx,
stopIdx=stopIdx,
recursive=recursive)
else:
# 'tif' or 'tif-stack'
loader.saveFromTif(dataPath,
outputDirPath,
ext=ext,
blockSize=blockSize,
startIdx=startIdx,
stopIdx=stopIdx,
overwrite=overwrite,
recursive=recursive)
def test_roundtripConvertToSeries(self):
imagepath = findSourceTreeDir("utils/data/fish/images")
images = ImagesLoader(self.sc).fromTif(imagepath)
strategy = SimpleBlockingStrategy.generateFromBlockSize(images, blockSize=76 * 20)
self._run_tst_roundtripConvertToSeries(images, strategy)
class TestImagesMeanByRegions(PySparkTestCase):
def setUp(self):
super(TestImagesMeanByRegions, self).setUp()
self.ary1 = array([[3, 5], [6, 8]], dtype='int32')
self.ary2 = array([[13, 15], [16, 18]], dtype='int32')
self.images = ImagesLoader(self.sc).fromArrays([self.ary1, self.ary2])
def __checkAttrPropagation(self, newImages, newDims):
assert_equals(newDims, newImages._dims.count)
assert_equals(self.images._nrecords, newImages._nrecords)
assert_equals(self.images._dtype, newImages._dtype)
def test_badMaskShapeThrowsValueError(self):
mask = array([[1]], dtype='int16')
assert_raises(ValueError, self.images.meanByRegions, mask)
def test_meanWithFloatMask(self):
mask = array([[1.0, 0.0], [0.0, 1.0]], dtype='float32')
regionMeanImages = self.images.meanByRegions(mask)
self.__checkAttrPropagation(regionMeanImages, (1, 1))
collected = regionMeanImages.collect()
assert_equals(2, len(collected))
assert_equals((1, 1), collected[0][1].shape)
# check keys
assert_equals(0, collected[0][0])
assert_equals(1, collected[1][0])
# check values
assert_equals(5, collected[0][1][0])
assert_equals(15, collected[1][1][0])
def test_meanWithIntMask(self):
mask = array([[1, 0], [2, 1]], dtype='uint8')
regionMeanImages = self.images.meanByRegions(mask)
self.__checkAttrPropagation(regionMeanImages, (1, 2))
collected = regionMeanImages.collect()
assert_equals(2, len(collected))
assert_equals((1, 2), collected[0][1].shape)
# check keys
assert_equals(0, collected[0][0])
assert_equals(1, collected[1][0])
# check values
assert_equals(5, collected[0][1].flat[0])
assert_equals(6, collected[0][1].flat[1])
assert_equals(15, collected[1][1].flat[0])
assert_equals(16, collected[1][1].flat[1])
def test_meanWithSingleRegionIndices(self):
indices = [[(1, 1), (0, 0)]] # one region with two indices
regionMeanImages = self.images.meanByRegions(indices)
self.__checkAttrPropagation(regionMeanImages, (1, 1))
collected = regionMeanImages.collect()
assert_equals(2, len(collected))
assert_equals((1, 1), collected[0][1].shape)
# check keys
assert_equals(0, collected[0][0])
assert_equals(1, collected[1][0])
# check values
assert_equals(5, collected[0][1][0])
assert_equals(15, collected[1][1][0])
def test_meanWithMultipleRegionIndices(self):
indices = [[(0, 0), (0, 1)],
[(0, 1), (1, 0)]] # two regions with two indices each
regionMeanImages = self.images.meanByRegions(indices)
self.__checkAttrPropagation(regionMeanImages, (1, 2))
collected = regionMeanImages.collect()
assert_equals(2, len(collected))
assert_equals((1, 2), collected[0][1].shape)
# check keys
assert_equals(0, collected[0][0])
assert_equals(1, collected[1][0])
# check values
assert_equals(4, collected[0][1].flat[0])
assert_equals(5, collected[0][1].flat[1])
assert_equals(14, collected[1][1].flat[0])
assert_equals(15, collected[1][1].flat[1])
def test_badIndexesThrowErrors(self):
indices = [[(0, 0), (-1, 0)]] # index too small (-1)
assert_raises(ValueError, self.images.meanByRegions, indices)
indices = [[(0, 0), (2, 0)]] # index too large (2)
assert_raises(ValueError, self.images.meanByRegions, indices)
indices = [[(0, 0), (0, )]] # too few indices
assert_raises(ValueError, self.images.meanByRegions, indices)
indices = [[(0, 0), (0, 1, 0)]] # too many indices
assert_raises(ValueError, self.images.meanByRegions, indices)
def test_meanWithSingleRegionIndices3D(self):
ary1 = array([[[3, 5, 3], [6, 8, 6]], [[3, 5, 3], [6, 8, 6]]],
dtype='int32')
ary2 = array(
[[[13, 15, 13], [16, 18, 16]], [[13, 15, 13], [16, 18, 16]]],
dtype='int32')
images = ImagesLoader(self.sc).fromArrays([ary1, ary2])
indices = [[(1, 1, 1), (0, 0, 0)]] # one region with two indices
regionMeanImages = images.meanByRegions(indices)
self.__checkAttrPropagation(regionMeanImages, (1, 1))
collected = regionMeanImages.collect()
# check values
assert_equals(5, collected[0][1][0])
assert_equals(15, collected[1][1][0])
def loadImagesAsSeries(self, dataPath, dims=None, inputFormat='stack', ext=None, dtype='int16',
blockSize="150M", blockSizeUnits="pixels", startIdx=None, stopIdx=None,
shuffle=True, recursive=False, nplanes=None, npartitions=None,
renumber=False, confFilename='conf.json'):
"""
Load Images data as Series data.
Parameters
----------
dataPath: string
Path to data files or directory, as either a local filesystem path or a URI.
May include a single '*' wildcard in the filename. Examples of valid dataPaths include
'local/directory/*.stack", "s3n:///my-s3-bucket/data/", or "file:///mnt/another/directory/".
dims: tuple of positive int, optional (required if inputFormat is 'stack')
Image dimensions. Binary stack data will be interpreted as a multidimensional array
with the given dimensions, and should be stored in row-major order (Fortran or Matlab convention),
where the first dimension changes most rapidly. For 'png' or 'tif' data dimensions
will be read from the image file headers.
inputFormat: str, optional, default = 'stack'
Expected format of the input data: 'stack', 'png', or 'tif'. 'stack' indicates flat binary stacks.
'png' or 'tif' indicate image formats. Page of a multipage tif file will be extend along
the third dimension. Separate files interpreted as distinct records, with ordering
given by lexicographic sorting of file names.
ext: string, optional, default = None
File extension, default will be "bin" if inputFormat=="stack", "tif" for inputFormat=='tif',
and 'png' for inputFormat=="png".
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string.
Ignored for 'tif' or 'png' (data will be inferred from image formats).
blockSize: string or positive int, optional, default "150M"
Requested size of blocks (e.g "64M", "512k", "2G"). If shuffle=True, can also be a
tuple of int specifying the number of pixels or splits per dimension. Indirectly
controls the number of Spark partitions, with one partition per block.
blockSizeUnits: string, either "pixels" or "splits", default "pixels"
Units for interpreting a tuple passed as blockSize when shuffle=True.
startIdx: nonnegative int, optional, default = None
Convenience parameters to read only a subset of input files. Uses python slice conventions
(zero-based indexing with exclusive final position). These parameters give the starting
and final index after lexicographic sorting.
stopIdx: nonnegative int, optional, default = None
See startIdx.
shuffle: boolean, optional, default = True
Controls whether the conversion from Images to Series formats will use of a Spark shuffle-based method.
recursive: boolean, optional, default = False
If true, will recursively descend directories rooted at dataPath, loading all files
in the tree with an appropriate extension.
nplanes: positive integer, optional, default = None
Subdivide individual image files. Every `nplanes` from each file will be considered a new record.
With nplanes=None (the default), a single file will be considered as representing a single record.
If the number of records per file is not the same across all files, then `renumber` should be set
to True to ensure consistent keys.
npartitions: positive int, optional, default = None
Specify number of partitions for the RDD, if unspecified will use 1 partition per image.
renumber: boolean, optional, default = False
Recalculate keys for records after images are loading. Only necessary if different files contain
different number of records (e.g. due to specifying nplanes). See Images.renumber().
confFilename : string, optional, default = 'conf.json'
Name of conf file if using to specify parameters for binary stack data
Returns
-------
data: thunder.rdds.Series
A Series object, wrapping an RDD, with (n-tuples of ints) : (numpy array) pairs.
Keys will be n-tuples of int, with n given by dimensionality of the images, and correspond
to indexes into the image arrays. Value will have length equal to the number of image files.
With each image contributing one point to this value array, with ordering given by
the lexicographic ordering of image file names.
"""
checkParams(inputFormat, ['stack', 'tif', 'tif-stack'])
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if shuffle:
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputFormat.lower() == 'stack':
images = loader.fromStack(dataPath, dims, dtype=dtype, ext=ext, startIdx=startIdx, stopIdx=stopIdx,
recursive=recursive, nplanes=nplanes, npartitions=npartitions,
confFilename=confFilename)
else:
# tif / tif stack
images = loader.fromTif(dataPath, ext=ext, startIdx=startIdx, stopIdx=stopIdx,
recursive=recursive, nplanes=nplanes, npartitions=npartitions)
if renumber:
images = images.renumber()
return images.toBlocks(blockSize, units=blockSizeUnits).toSeries()
else:
from thunder.rdds.fileio.seriesloader import SeriesLoader
if nplanes is not None:
raise NotImplementedError("nplanes is not supported with shuffle=False")
if npartitions is not None:
raise NotImplementedError("npartitions is not supported with shuffle=False")
if renumber:
raise NotImplementedError("renumber is not supported with shuffle=False")
loader = SeriesLoader(self._sc)
if inputFormat.lower() == 'stack':
return loader.fromStack(dataPath, dims, ext=ext, dtype=dtype, blockSize=blockSize,
startIdx=startIdx, stopIdx=stopIdx, recursive=recursive)
else:
# tif / tif stack
return loader.fromTif(dataPath, ext=ext, blockSize=blockSize,
startIdx=startIdx, stopIdx=stopIdx, recursive=recursive)
def convertImagesToSeries(self, dataPath, outputDirPath, dims=None, inputFormat='stack', ext=None,
dtype='int16', blockSize="150M", blockSizeUnits="pixels", startIdx=None, stopIdx=None,
shuffle=True, overwrite=False, recursive=False, nplanes=None, npartitions=None,
renumber=False, confFilename='conf.json'):
"""
Write out Images data as Series data, saved in a flat binary format.
The resulting files may subsequently be read in using ThunderContext.loadSeries().
Loading Series data directly will likely be faster than converting image data
to a Series object through loadImagesAsSeries().
Parameters
----------
dataPath: string
Path to data files or directory, as either a local filesystem path or a URI.
May include a single '*' wildcard in the filename. Examples of valid dataPaths include
'local/directory/*.stack", "s3n:///my-s3-bucket/data/", or "file:///mnt/another/directory/".
outputDirPath: string
Path to directory to write Series file output. May be either a path on the local file system
or a URI-like format, such as "local/directory", "s3n:///my-s3-bucket/data/",
or "file:///mnt/another/directory/". If the directory exists and 'overwrite' is True,
the existing directory and all its contents will be deleted and overwritten.
dims: tuple of positive int, optional (required if inputFormat is 'stack')
Image dimensions. Binary stack data will be interpreted as a multidimensional array
with the given dimensions, and should be stored in row-major order (Fortran or Matlab convention),
where the first dimension changes most rapidly. For 'png' or 'tif' data dimensions
will be read from the image file headers.
inputFormat: str, optional, default = 'stack'
Expected format of the input data: 'stack', 'png', or 'tif'. 'stack' indicates flat binary stacks.
'png' or 'tif' indicate image formats. Page of a multipage tif file will be extend along
the third dimension. Separate files interpreted as distinct records, with ordering
given by lexicographic sorting of file names.
ext: string, optional, default = None
File extension, default will be "bin" if inputFormat=="stack", "tif" for inputFormat=='tif',
and 'png' for inputFormat=="png".
dtype: string or numpy dtype. optional, default 'int16'
Data type of the image files to be loaded, specified as a numpy "dtype" string.
Ignored for 'tif' or 'png' (data will be inferred from image formats).
blockSize: string or positive int, optional, default "150M"
Requested size of blocks (e.g "64M", "512k", "2G"). If shuffle=True, can also be a
tuple of int specifying the number of pixels or splits per dimension. Indirectly
controls the number of Spark partitions, with one partition per block.
blockSizeUnits: string, either "pixels" or "splits", default "pixels"
Units for interpreting a tuple passed as blockSize when shuffle=True.
startIdx: nonnegative int, optional, default = None
Convenience parameters to read only a subset of input files. Uses python slice conventions
(zero-based indexing with exclusive final position). These parameters give the starting
and final index after lexicographic sorting.
stopIdx: nonnegative int, optional, default = None
See startIdx.
shuffle: boolean, optional, default = True
Controls whether the conversion from Images to Series formats will use of a Spark shuffle-based method.
overwrite: boolean, optional, default False
If true, the directory specified by outputDirPath will be deleted (recursively) if it
already exists. (Use with caution.)
recursive: boolean, optional, default = False
If true, will recursively descend directories rooted at dataPath, loading all files
in the tree with an appropriate extension.
nplanes: positive integer, optional, default = None
Subdivide individual image files. Every `nplanes` from each file will be considered a new record.
With nplanes=None (the default), a single file will be considered as representing a single record.
If the number of records per file is not the same across all files, then `renumber` should be set
to True to ensure consistent keys.
npartitions: positive int, optional, default = None
Specify number of partitions for the RDD, if unspecified will use 1 partition per image.
renumber: boolean, optional, default = False
Recalculate keys for records after images are loading. Only necessary if different files contain
different number of records (e.g. due to specifying nplanes). See Images.renumber().
confFilename : string, optional, default = 'conf.json'
Name of conf file if using to specify parameters for binary stack data
"""
checkParams(inputFormat, ['stack', 'tif', 'tif-stack'])
if not overwrite:
raiseErrorIfPathExists(outputDirPath, awsCredentialsOverride=self._credentials)
overwrite = True # prevent additional downstream checks for this path
if not ext:
ext = DEFAULT_EXTENSIONS.get(inputFormat.lower(), None)
if shuffle:
from thunder.rdds.fileio.imagesloader import ImagesLoader
loader = ImagesLoader(self._sc)
if inputFormat.lower() == 'stack':
images = loader.fromStack(dataPath, dims, ext=ext, dtype=dtype, startIdx=startIdx, stopIdx=stopIdx,
recursive=recursive, nplanes=nplanes, npartitions=npartitions,
confFilename=confFilename)
else:
# 'tif' or 'tif-stack'
images = loader.fromTif(dataPath, ext=ext, startIdx=startIdx, stopIdx=stopIdx,
recursive=recursive, nplanes=nplanes, npartitions=npartitions)
if renumber:
images = images.renumber()
images.toBlocks(blockSize, units=blockSizeUnits).saveAsBinarySeries(outputDirPath, overwrite=overwrite)
else:
from thunder.rdds.fileio.seriesloader import SeriesLoader
if nplanes is not None:
raise NotImplementedError("nplanes is not supported with shuffle=False")
if npartitions is not None:
raise NotImplementedError("npartitions is not supported with shuffle=False")
loader = SeriesLoader(self._sc)
if inputFormat.lower() == 'stack':
loader.saveFromStack(dataPath, outputDirPath, dims, ext=ext, dtype=dtype,
blockSize=blockSize, overwrite=overwrite, startIdx=startIdx,
stopIdx=stopIdx, recursive=recursive)
else:
# 'tif' or 'tif-stack'
loader.saveFromTif(dataPath, outputDirPath, ext=ext, blockSize=blockSize,
startIdx=startIdx, stopIdx=stopIdx, overwrite=overwrite,
recursive=recursive)
def roundTrip(images, dtype):
outdir = os.path.join(self.outputdir, "binary-images-" + dtype)
images.astype(dtype).saveAsBinaryImages(outdir)
newimages = ImagesLoader(self.sc).fromStack(outdir, ext='bin')
array_equal(images.first()[1], newimages.first()[1])
