def read(infile, fmt="tab", into=None, sample_id=None, meta=None, **kwargs):
"""Read tabular data from a file or stream into a genome object.
Supported formats: see `READERS`
If a format supports multiple samples, return the sample specified by
`sample_id`, or if unspecified, return the first sample and warn if there
were other samples present in the file.
Parameters
----------
infile : handle or string
Filename or opened file-like object to read.
fmt : string
File format.
into : class
GenomicArray class or subclass to instantiate, overriding the
default for the target file format.
sample_id : string
Sample identifier.
meta : dict
Metadata, as arbitrary key-value pairs.
**kwargs :
Additional keyword arguments to the format-specific reader function.
Returns
-------
GenomicArray or subclass
The data from the given file instantiated as `into`, if specified, or
the default base class for the given file format (usually GenomicArray).
"""
from cnvlib.core import fbase
if fmt == 'auto':
return read_auto(infile)
elif fmt in READERS:
reader, suggest_into = READERS[fmt]
else:
raise ValueError("Unknown format: %s" % fmt)
if meta is None:
meta = {}
if "sample_id" not in meta:
if sample_id:
meta["sample_id"] = sample_id
elif isinstance(infile, basestring):
meta["sample_id"] = fbase(infile)
elif hasattr(infile, "name"):
meta["sample_id"] = fbase(infile.name)
else:
# meta["sample_id"] = "<unknown>"
pass
if "filename" not in meta:
if isinstance(infile, basestring):
meta["filename"] = infile
elif hasattr(infile, "name"):
meta["filename"] = infile.name
if fmt in ("seg", "vcf") and sample_id is not None:
# Multi-sample formats: choose one sample
kwargs["sample_id"] = sample_id
try:
dframe = reader(infile, **kwargs)
except pd.io.common.EmptyDataError:
# File is blank/empty, most likely
logging.info("Blank %s file?: %s", fmt, infile)
dframe = []
if fmt == "vcf":
from cnvlib.vary import VariantArray as VA
suggest_into = VA
result = (into or suggest_into)(dframe, meta)
result.sort_columns()
result.sort()
return result
def read(infile, fmt="tab", into=None, sample_id=None, meta=None, **kwargs):
"""Read tabular data from a file or stream into a genome object.
Supported formats: see `READERS`
If a format supports multiple samples, return the sample specified by
`sample_id`, or if unspecified, return the first sample and warn if there
were other samples present in the file.
Parameters
----------
infile : handle or string
Filename or opened file-like object to read.
fmt : string
File format.
into : class
GenomicArray class or subclass to instantiate, overriding the
default for the target file format.
sample_id : string
Sample identifier.
meta : dict
Metadata, as arbitrary key-value pairs.
**kwargs :
Additional keyword arguments to the format-specific reader function.
Returns
-------
GenomicArray or subclass
The data from the given file instantiated as `into`, if specified, or
the default base class for the given file format (usually GenomicArray).
"""
from cnvlib.core import fbase
if fmt == 'auto':
return read_auto(infile)
elif fmt in READERS:
reader, suggest_into = READERS[fmt]
else:
raise ValueError("Unknown format: %s" % fmt)
if meta is None:
meta = {}
if "sample_id" not in meta:
if sample_id:
meta["sample_id"] = sample_id
else:
fname = get_filename(infile)
if fname:
meta["sample_id"] = fbase(fname)
if "filename" not in meta:
fname = get_filename(infile)
if fname:
meta["filename"] = infile
if fmt in ("seg", "vcf") and sample_id is not None:
# Multi-sample formats: choose one sample
kwargs["sample_id"] = sample_id
try:
dframe = reader(infile, **kwargs)
except pd.io.common.EmptyDataError:
# File is blank/empty, most likely
logging.info("Blank %s file?: %s", fmt, infile)
dframe = []
if fmt == "vcf":
from cnvlib.vary import VariantArray as VA
suggest_into = VA
result = (into or suggest_into)(dframe, meta)
result.sort_columns()
result.sort()
return result