def write(filename, kerneldata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`KernelData` to KernelHdf5 format and returns the :class:`.KernelHdf5`.
:param filename: the name of the file to create
:type filename: string
:param kerneldata: The in-memory data that should be written to disk.
:type kerneldata: :class:`KernelData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param col_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
:rtype: :class:`.KernelHdf5`
>>> from pysnptools.snpreader import Bed
>>> from pysnptools.standardizer import Unit
>>> import pysnptools.util as pstutil
>>> from pysnptools.util import example_file # Download and return local file name
>>> bed_file = example_file('pysnptools/examples/toydata.5chrom.*','*.bed')
>>> kerneldata = Bed(bed_file,count_A1=False).read_kernel(Unit()) # Create a kernel from the data in the Bed file
>>> pstutil.create_directory_if_necessary("tempdir/toydata.kernel.hdf5")
>>> KernelHdf5.write("tempdir/toydata.kernel.hdf5",kerneldata) # Write data in KernelHdf5 format
KernelHdf5('tempdir/toydata.kernel.hdf5')
"""
PstHdf5.write(filename,kerneldata,hdf5_dtype=hdf5_dtype,col_major=sid_major)
return KernelHdf5(filename)
def write(filename, snpdata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`SnpData` to SnpHdf5 format and return a the :class:`.SnpHdf5`.
:param filename: the name of the file to create
:type filename: string
:param snpdata: The in-memory data that should be written to disk.
:type snpdata: :class:`SnpData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param sid_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type sid_major: bool
:rtype: :class:`.SnpHdf5`
>>> from pysnptools.snpreader import SnpHdf5, Bed
>>> import pysnptools.util as pstutil
>>> from pysnptools.util import example_file # Download and return local file name
>>> bedfile = example_file("pysnptools/examples/toydata.5chrom.*","*.bed")
>>> snpdata = Bed(bedfile,count_A1=False)[:,:10].read() # Read first 10 snps from Bed format
>>> pstutil.create_directory_if_necessary("tempdir/toydata10.snp.hdf5")
>>> SnpHdf5.write("tempdir/toydata10.snp.hdf5",snpdata) # Write data in SnpHdf5 format
SnpHdf5('tempdir/toydata10.snp.hdf5')
"""
PstHdf5.write(filename,
snpdata,
hdf5_dtype=hdf5_dtype,
col_major=sid_major)
return SnpHdf5(filename)
def write(filename, distdata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`DistData` to DistHdf5 format and return a the :class:`.DistHdf5`.
:param filename: the name of the file to create
:type filename: string
:param distdata: The in-memory data that should be written to disk.
:type distdata: :class:`DistData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param sid_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
:rtype: :class:`.DistHdf5`
>>> from pysnptools.distreader import DistHdf5, Bgen
>>> import pysnptools.util as pstutil
>>> from pysnptools.util import example_file # Download and return local file name
>>> bgen_file = example_file("pysnptools/examples/2500x100.bgen")
>>> distdata = Bgen(bgen_file)[:,:10].read() # Read first 10 snps from BGEN format
>>> pstutil.create_directory_if_necessary("tempdir/toydata10.dist.hdf5")
>>> DistHdf5.write("tempdir/toydata10.dist.hdf5",distdata) # Write data in DistHdf5 format
DistHdf5('tempdir/toydata10.dist.hdf5')
"""
PstHdf5.write(filename,
distdata,
hdf5_dtype=hdf5_dtype,
col_major=sid_major)
return DistHdf5(filename)
def write(filename, kerneldata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`KernelData` to KernelHdf5 format.
:param filename: the name of the file to create
:type filename: string
:param kerneldata: The in-memory data that should be written to disk.
:type kerneldata: :class:`KernelData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param col_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
"""
# >>> from pysnptools.snpreader import Bed
# >>> from pysnptools.standardizer import Unit
# >>> import pysnptools.util as pstutil
# >>> kerneldata = Bed('pysnptools/examples/toydata.bed',count_A1=False).read_kernel(Unit()) # Create a kernel from the data in the Bed file
# >>> pstutil.create_directory_if_necessary(b"tempdir/toydata.kernel.hdf5")
# >>> KernelHdf5.write("tempdir/toydata.kernel.hdf5",kerneldata) # Write data in KernelHdf5 format
PstHdf5.write(filename,kerneldata,hdf5_dtype=hdf5_dtype,col_major=sid_major)
def write(filename, snpdata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`SnpData` to SnpHdf5 format.
:param filename: the name of the file to create
:type filename: string
:param snpdata: The in-memory data that should be written to disk.
:type snpdata: :class:`SnpData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param col_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
>>> from pysnptools.snpreader import SnpHdf5, Bed
>>> import pysnptools.util as pstutil
>>> snpdata = Bed('../examples/toydata.bed')[:,:10].read() # Read first 10 snps from Bed format
>>> pstutil.create_directory_if_necessary("tempdir/toydata10.snp.hdf5")
>>> SnpHdf5.write("tempdir/toydata10.snp.hdf5",snpdata) # Write data in SnpHdf5 format
"""
PstHdf5.write(filename,snpdata,hdf5_dtype=hdf5_dtype,col_major=sid_major)
def test_flush(self):
previous_wd = os.getcwd()
os.chdir(os.path.dirname(os.path.realpath(__file__)))
reader = PstHdf5('../examples/toydata.kernel.hdf5')
val1 = reader[0, 0].read()
reader.flush()
val2 = reader[0, 0].read()
assert val1 == val2
def write(filename, snpdata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`SnpData` to SnpHdf5 format.
:param filename: the name of the file to create
:type filename: string
:param snpdata: The in-memory data that should be written to disk.
:type snpdata: :class:`SnpData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param col_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
>>> from pysnptools.snpreader import SnpHdf5, Bed
>>> import pysnptools.util as pstutil
"""
PstHdf5.write(filename,
snpdata,
hdf5_dtype=hdf5_dtype,
col_major=sid_major)
def write(filename, kerneldata, hdf5_dtype=None, sid_major=True):
"""Writes a :class:`KernelData` to KernelHdf5 format.
:param filename: the name of the file to create
:type filename: string
:param kerneldata: The in-memory data that should be written to disk.
:type kerneldata: :class:`KernelData`
:param hdf5_dtype: None (use the .val's dtype) or a Hdf5 dtype, e.g. 'f8','f4',etc.
:type hdf5_dtype: string
:param col_major: Tells if vals should be stored on disk in sid_major (default) or iid_major format.
:type col_major: bool
>>> from pysnptools.snpreader import Bed
>>> from pysnptools.standardizer import Unit
>>> import pysnptools.util as pstutil
>>> kerneldata = Bed('../examples/toydata.bed').read_kernel(Unit()) # Create a kernel from the data in the Bed file
>>> pstutil.create_directory_if_necessary("tempdir/toydata.kernel.hdf5")
>>> KernelHdf5.write("tempdir/toydata.kernel.hdf5",kerneldata) # Write data in KernelHdf5 format
"""
PstHdf5.write(filename,kerneldata,hdf5_dtype=hdf5_dtype,col_major=sid_major)
def test_respect_read_inputs(self):
from pysnptools.pstreader import _MergeRows, _MergeCols
previous_wd = os.getcwd()
os.chdir(os.path.dirname(os.path.realpath(__file__)))
for pstreader in [
PstNpz('../examples/toydata10.snp.npz'),
_MergeRows([
PstHdf5('../examples/toydata.snpmajor.snp.hdf5')
[:5, :].read(),
PstHdf5('../examples/toydata.snpmajor.snp.hdf5')[
5:, :].read()
]),
_MergeCols([
PstHdf5(
'../examples/toydata.snpmajor.snp.hdf5')[:, :5].read(),
PstHdf5('../examples/toydata.snpmajor.snp.hdf5')
[:, 5:].read()
]),
PstHdf5('../examples/toydata.snpmajor.snp.hdf5')[::2, ::2],
PstHdf5('../examples/toydata.snpmajor.dist.hdf5').read(),
PstHdf5('../examples/toydata.kernel.hdf5'),
PstMemMap('../examples/tiny.pst.memmap')
]:
logging.info(str(pstreader))
for order in ['F', 'C', 'A']:
for dtype in [np.float32, np.float64]:
for force_python_only in [True, False]:
for view_ok in [True, False]:
val = pstreader.read(
order=order,
dtype=dtype,
force_python_only=force_python_only,
view_ok=view_ok).val
has_right_order = order == "A" or (
order == "C" and val.flags["C_CONTIGUOUS"]
) or (order == "F" and val.flags["F_CONTIGUOUS"])
if hasattr(pstreader, 'val') and not view_ok:
assert pstreader.val is not val
if (hasattr(pstreader, 'val') and view_ok
and pstreader.val is not val and
(order == 'A' or
(order == 'F'
and pstreader.val.flags['F_CONTIGUOUS']) or
(order == 'C'
and pstreader.val.flags['C_CONTIGUOUS']))
and (dtype is None
or pstreader.val.dtype == dtype)):
logging.info(
"{0} could have read a view, but didn't".
format(pstreader))
assert val.dtype == dtype and has_right_order
os.chdir(previous_wd)