def totif(images, path, prefix="image", overwrite=False, credentials=None):
"""
Write out TIF files for 2d image data.
See also
--------
thunder.data.images.totif
"""
dims = images.dims
if not len(dims) in [2, 3]:
raise ValueError("Only 2D or 3D images can be exported to tif, "
"images are %d-dimensional." % len(dims))
from scipy.misc import imsave
from io import BytesIO
from thunder.writers import get_parallel_writer
def tobuffer(kv):
key, img = kv
fname = prefix + "-" + "%05d.tif" % int(key)
bytebuf = BytesIO()
imsave(bytebuf, img, format='TIFF')
return fname, bytebuf.getvalue()
writer = get_parallel_writer(path)(path,
overwrite=overwrite,
credentials=credentials)
images.foreach(lambda x: writer.write(tobuffer(x)))
def totif(images, path, prefix="image", overwrite=False, credentials=None):
"""
Write out TIF files for 2d image data.
See also
--------
thunder.data.images.totif
"""
dims = images.dims
if not len(dims) in [2, 3]:
raise ValueError("Only 2D or 3D images can be exported to tif, "
"images are %d-dimensional." % len(dims))
from scipy.misc import imsave
from io import BytesIO
from thunder.writers import get_parallel_writer
def tobuffer(kv):
key, img = kv
fname = prefix+"-"+"%05d.tif" % int(key)
bytebuf = BytesIO()
imsave(bytebuf, img, format='TIFF')
return fname, bytebuf.getvalue()
writer = get_parallel_writer(path)(path, overwrite=overwrite, credentials=credentials)
images.foreach(lambda x: writer.write(tobuffer(x)))
def tobinary(images, path, prefix="image", overwrite=False, credentials=None):
"""
Write out images as binary files.
See also
--------
thunder.data.images.tobinary
"""
from thunder.writers import get_parallel_writer
def tobuffer(kv):
key, img = kv
fname = prefix + "-" + "%05d.bin" % int(key)
return fname, img.copy()
writer = get_parallel_writer(path)(path, overwrite=overwrite, credentials=credentials)
images.foreach(lambda x: writer.write(tobuffer(x)))
config(path, list(images.dims), images.dtype, overwrite=overwrite)
def tobinary(images, path, prefix="image", overwrite=False, credentials=None):
"""
Write out images as binary files.
See also
--------
thunder.data.images.tobinary
"""
from thunder.writers import get_parallel_writer
def tobuffer(kv):
key, img = kv
fname = prefix + "-" + "%05d.bin" % int(key)
return fname, img.copy()
writer = get_parallel_writer(path)(path,
overwrite=overwrite,
credentials=credentials)
images.foreach(lambda x: writer.write(tobuffer(x)))
config(path, list(images.dims), images.dtype, overwrite=overwrite)
def tobinary(series, path, prefix='series', overwrite=False, credentials=None):
"""
Writes out data to binary format.
Parameters
----------
series : Series
The data to write
path : string path or URI to directory to be created
Output files will be written underneath path.
Directory will be created as a result of this call.
prefix : str, optional, default = 'series'
String prefix for files.
overwrite : bool
If true, path and all its contents will be deleted and
recreated as partof this call.
"""
from six import BytesIO
from thunder.utils import check_path
from thunder.writers import get_parallel_writer
if not overwrite:
check_path(path, credentials=credentials)
overwrite = True
def tobuffer(kv):
firstkey = None
buf = BytesIO()
for k, v in kv:
if firstkey is None:
firstkey = k
buf.write(v.tostring())
val = buf.getvalue()
buf.close()
if firstkey is None:
return iter([])
else:
label = prefix + '-' + getlabel(firstkey) + ".bin"
return iter([(label, val)])
writer = get_parallel_writer(path)(path,
overwrite=overwrite,
credentials=credentials)
if series.mode == 'spark':
binary = series.values.tordd().sortByKey().mapPartitions(tobuffer)
binary.foreach(writer.write)
else:
basedims = [series.shape[d] for d in series.baseaxes]
def split(k):
ind = unravel_index(k, basedims)
return ind, series.values[ind]
buf = tobuffer([split(i) for i in range(prod(basedims))])
[writer.write(b) for b in buf]
shape = series.shape
dtype = series.dtype
write_config(path,
shape=shape,
dtype=dtype,
overwrite=overwrite,
credentials=credentials)
def tobinary(series, path, prefix='series', overwrite=False, credentials=None):
"""
Writes out data to binary format.
Parameters
----------
series : Series
The data to write
path : string path or URI to directory to be created
Output files will be written underneath path.
Directory will be created as a result of this call.
prefix : str, optional, default = 'series'
String prefix for files.
overwrite : bool
If true, path and all its contents will be deleted and
recreated as partof this call.
"""
from six import BytesIO
from thunder.utils import check_path
from thunder.writers import get_parallel_writer
if not overwrite:
check_path(path, credentials=credentials)
overwrite = True
def tobuffer(kv):
firstkey = None
buf = BytesIO()
for k, v in kv:
if firstkey is None:
firstkey = k
buf.write(v.tostring())
val = buf.getvalue()
buf.close()
if firstkey is None:
return iter([])
else:
label = prefix + '-' + getlabel(firstkey) + ".bin"
return iter([(label, val)])
writer = get_parallel_writer(path)(path, overwrite=overwrite, credentials=credentials)
if series.mode == 'spark':
binary = series.values.tordd().sortByKey().mapPartitions(tobuffer)
binary.foreach(writer.write)
else:
basedims = [series.shape[d] for d in series.baseaxes]
def split(k):
ind = unravel_index(k, basedims)
return ind, series.values[ind]
buf = tobuffer([split(i) for i in range(prod(basedims))])
[writer.write(b) for b in buf]
shape = series.shape
dtype = series.dtype
write_config(path, shape=shape, dtype=dtype, overwrite=overwrite, credentials=credentials)