def write(self, metadata, samples, filename):
'''
samples - list of samples
filename - name of output file
'''
separator = "\t"
with open(filename, "w") as out:
# metadata - labels
out.write("EVENTS%s" % (separator))
for label in metadata.labels:
out.write("%s%s" % (label, separator))
out.write("class\n")
# actual data
i = 2
for sample in samples:
# description in the 1st column
out.write("%s%s" % (sample.description, separator))
# value
for value in sample.features:
if is_na(value):
out.write("?%s" % (separator))
elif isinstance(value, numbers.Integral):
int_value = int(value)
out.write("%d%s" % (int_value, separator))
elif isinstance(value, numbers.Real):
out.write("%2.2f%s" % (value, separator))
else:
raise UnexpectedDataError
# class
out.write("%s\n"
% (sample.classification))
def write(self, metadata, samples, filename):
'''
samples - list of samples
filename - name of output file
'''
desc_filename = filename + "-description"
separator = " "
with open(filename, "w") as out, open(desc_filename, "w") as desc:
# header
out.write("Number of examples %d\n" % (len(samples)))
try:
samples_number = len(samples[0].features)
except IndexError:
samples_number = 0
out.write("Number of inputs %d\n" % (samples_number))
# metadata - types
out.write("n%s" % (separator))
for data_type in metadata.types:
out.write("%s%s" % (data_type, separator))
out.write("o\n")
# metadata - labels
out.write("no%s" % (separator))
for label in metadata.labels:
out.write("%s%s" % (label, separator))
out.write("class\n")
# actual data
i = 1
for sample in samples:
# ordinal number
out.write("%d%s" % (i, separator))
# description in a separate file
desc.write("%d%s%s\n" %
(i, separator, sample.description.lower()))
i = i + 1
# value
for value in sample.features:
if is_na(value):
out.write("?%s" % (separator))
elif isinstance(value, numbers.Integral):
int_value = int(value)
out.write("%d%s" % (int_value, separator))
elif isinstance(value, numbers.Real):
out.write("%2.2f%s" % (value, separator))
else:
raise UnexpectedDataError
# class
out.write("%d\n" %
(self.translate_to_numeric(sample.classification)))
def write(self, metadata, samples, filename):
'''
samples - list of samples
filename - name of output file
'''
desc_filename = filename+"-description"
separator = " "
with open(filename, "w") as out, open(desc_filename, "w") as desc:
# header
out.write("Number of examples %d\n" % (len(samples)))
try:
samples_number = len(samples[0].features)
except IndexError:
samples_number = 0
out.write("Number of inputs %d\n" % (samples_number))
# metadata - types
out.write("n%s" % (separator))
for data_type in metadata.types:
out.write("%s%s" % (data_type, separator))
out.write("o\n")
# metadata - labels
out.write("no%s" % (separator))
for label in metadata.labels:
out.write("%s%s" % (label, separator))
out.write("class\n")
# actual data
i = 1
for sample in samples:
# ordinal number
out.write("%d%s" % (i, separator))
# description in a separate file
desc.write("%d%s%s\n" % (i, separator, sample.description.lower()))
i = i + 1
# value
for value in sample.features:
if is_na(value):
out.write("?%s" % (separator))
elif isinstance(value, numbers.Integral):
int_value = int(value)
out.write("%d%s" % (int_value, separator))
elif isinstance(value, numbers.Real):
out.write("%2.2f%s" % (value, separator))
else:
raise UnexpectedDataError
# class
out.write("%d\n"
% (self.translate_to_numeric(sample.classification)))
