def _df_to_libffm(self, out_file_or_y=None, y=None):
out_file = out_file_or_y if type(out_file_or_y) is str else None
if y is None and out_file_or_y is not None and out_file is None:
y = out_file_or_y
if hasattr(y, 'values'): y = y.values
lines = []
outfile = None
if out_file is not None: outfile = utils.get_write_file_stream(out_file)
categoricals = self.categoricals() + self.indexes() + self.binaries()
if len(categoricals) > 0: raise Exception('categoricals not currently supported')
numericals = self.numericals()
for idx, row in enumerate(utils.chunked_iterator(self)):
label = '0' if y is None or idx >= len(y) else str(int(y[idx]))
new_line = [label]
for col_idx, c in enumerate(numericals):
cis = str(col_idx)
new_line.append(cis + ':0:' + `0 if row[c] == 0 else row[c]`)
line = ' '.join(new_line)
if outfile is not None: outfile.write(line + '\n')
else: lines.append(line)
if outfile:
outfile.close()
return None
else: return lines
def impl(outfile):
def add_cols(new_line, columns, is_numerical):
if len(columns) == 0: return
if tag_feature_sets: new_line.append('|' + ('n' if is_numerical else 'c'))
for c in columns:
val = row[c]
if val == 0: continue
if not is_numerical:
name = c + '_' + str(val)
if output_categorical_value: line = get_col_index(name) + ':1'
else:
line = get_col_index(name)
else:
line = get_col_index(c) + ':' + str(val)
new_line.append(line)
lines = []
for idx, row in enumerate(utils.chunked_iterator(df)):
label = '1.0' if y is None or idx >= len(y) else str(float(y[idx]))
if convert_zero_ys and label == '0.0': label = '-1.0'
if weights is not None and idx < len(weights):
w = weights[idx]
if w != 1: label += ' ' + `w`
label += ' \'' + `idx`
new_line = []
add_cols(new_line, df.numericals(), True)
add_cols(new_line, df.categoricals() + df.indexes() + df.binaries(), False)
if sort_feature_indexes:
new_line = sorted(new_line, key=lambda v: int(v.split(':')[0]))
line = ' '.join([label] + new_line)
if outfile: outfile.write(line + '\n')
else: lines.append(line)
return lines
def _anonymize_by_obj(self, obj_anonymizer, qs):
bar = pyprind.ProgBar(qs.count(),
title='Anonymize model {}'.format(
self._get_full_model_name(qs.model)),
stream=ProgressBarStream(self.stdout))
for obj in chunked_iterator(qs, obj_anonymizer.chunk_size):
obj_anonymizer().anonymize_obj(obj)
bar.update()
