def __init__(self, **keywords):
"""
dataMatrixDtype could be a compound type:
http://docs.scipy.org/doc/numpy/reference/arrays.dtypes.html
http://docs.scipy.org/doc/numpy/reference/generated/numpy.dtype.html
#A record data type containing a 16-character string (in field name)
#and a sub-array of two 64-bit floating-point number (in field grades):
dt = numpy.dtype([('name', numpy.str_, 16), ('grades', numpy.float64, (2,))])
my_dtype = numpy.dtype([('field1', 'i'), ('field2', 'f'), ('field3', varLenStrType)])
#Using array-protocol type strings:
#each number below is counting bytes, not bits
>>> numpy.dtype([('a','f8'),('b','S10')])
dtype([('a', '<f8'), ('b', '|S10')])
#Using tuples. int is a fixed type, 3 the field's shape. void is a flexible type, here of size 10:
numpy.dtype([('hello',(numpy.int,3)),('world',numpy.void,10)])
#Using dictionaries. Two fields named 'gender' and 'age':
numpy.dtype({'names':['gender','age'], 'formats':['S1',numpy.uint8]})
#Offsets in bytes, here 0 and 25:
numpy.dtype({'surname':('S25',0),'age':(numpy.uint8,25)})
"""
self.ad = ProcessOptions.process_function_arguments(
keywords,
self.option_default_dict,
error_doc=self.__doc__,
class_to_have_attr=self)
self.dataMatrixDSName = "dataMatrix"
self.rowIDListDSName = "rowIDList"
self.colIDListDSName = "colIDList"
if not self.newGroup:
self._readInData()
else:
self._createDatasetSkeletonForOneGroup(h5Group=self.h5Group,
dtype=self.dataMatrixDtype)
self.newWrite = True
#a flag used to control whether it's first time to write stuff (first time=set whole matrix)
self.rowIndexCursor = 0
def __init__(self, path=None, **keywords):
self.ad = ProcessOptions.process_function_arguments(keywords,
self.option_default_dict, error_doc=self.__doc__,
class_to_have_attr=self)
if not self.path:
self.path = path
if self.path and self.file_handle is None:
self.file_handle = utils.openGzipFile(self.path, mode=self.mode)
#2013.05.03 for easy access
self.filename = self.path
self.csvFile = None
self.isRealCSV = False
if self.mode=='r': #reading mode
if self.delimiter is None:
self.delimiter = figureOutDelimiter(self.file_handle)
if self.delimiter=='\t' or self.delimiter==',':
self.csvFile = csv.reader(self.file_handle, delimiter=self.delimiter)
self.isRealCSV = True
else:
self.csvFile = self.file_handle
self.isRealCSV = False
else: #writing mode
if not self.delimiter:
self.delimiter = '\t'
self.csvFile = csv.writer(self.file_handle, delimiter=self.delimiter)
self.isRealCSV = True
#else:
# self.csvFile = self.file_handle
# self.isRealCSV = False
self.col_name2index = None
self._row = None # store the current row being read
self.headerPattern = re.compile(r'^[a-zA-Z]')
#default header pattern, line beginned with letter
self.commentPattern = re.compile(r'^#') #default, beginned with #
self.comment_row_list = []
def __init__(self, inputFname=None, **keywords):
self.ad = ProcessOptions.process_function_arguments(keywords,
self.option_default_dict, error_doc=self.__doc__, \
class_to_have_attr=self)
if not self.inputFname:
self.inputFname = inputFname
self.header = None
self.combinedColIDList = None #same as header
self.combinedColID2ColIndex = None
self.hdf5File = h5py.File(self.inputFname, self.mode)
self.tableObjectList = []
self.tablePath2Index = {}
if self.mode == 'r':
self._readInData()
elif self.mode == 'w':
self.createNewTable(tableName=self.tableName,
dtype=self.dtype,
rowDefinition=self.rowDefinition)
self.rowIndexCursor = 0 #2012.11.16 for iteration
def __init__(self, **keywords):
"""
2012.9.5 set default minDepth=0
2011-9-27
"""
self.ad = ProcessOptions.process_function_arguments(keywords,
self.option_default_dict, error_doc=self.__doc__, \
class_to_have_attr=self)
self.header = None
# the list of column headers (the header line starting by #CHROM)
self.headerWithoutHash = None
# same as self.header except, instead of "#CHROM", it is "CHROM".
self.sample_id_ls = []
self.sample_id2index = {}
#the index is the index of its column in the genotype_call_matrix
self.locus_id_ls = []
self.locus_id2row_index = {}
self.locus_id2data = {}
self.genotype_call_matrix = []
self.col_name2index = {} #column index in file
self.col_index_individual_name_ls = None
self.individual_name2col_index = {}
#not the matrix column, the column in input file
self.metaInfoLs = []
# anything before the "#CHROM" line. each entry is a raw line content, including '\n'
self.sampleIDHeader = []
# a list of sample column headers (from sampleStartingColumn)
self.inf = None
self.reader = None
self._initializeInput(self.inputFname)
self.outf = None
self.writer = None
self._initializeOutput(self.outputFname)
raise
else:
if key not in dc:
dc[key] = []
if len(valueColumnIndexList)>1:
value = valueList
else:
value = valueList[0]
if keyUniqueInInputFile:
dc[key] = value
else:
dc[key].append(value)
sys.stderr.write("%s unique pairs from %s rows.\n"%(len(dc), counter))
return dc
def run(self):
"""
"""
if self.debug:
import pdb
pdb.set_trace()
if __name__ == '__main__':
main_class = MatrixFile
po = ProcessOptions(sys.argv, main_class.option_default_dict, error_doc=main_class.__doc__)
instance = main_class(**po.long_option2value)
instance.run()