def _simple_parser(lines, colNames=None, header=0, index_col=0,
na_values=None, date_parser=None, parse_dates=True):
"""
Workhorse function for processing nested list into DataFrame
Should be replaced by np.genfromtxt eventually?
"""
if header is not None:
columns = []
for i, c in enumerate(lines[header]):
if c == '':
columns.append('Unnamed: %d' % i)
else:
columns.append(c)
content = lines[header+1:]
counts = {}
for i, col in enumerate(columns):
cur_count = counts.get(col, 0)
if cur_count > 0:
columns[i] = '%s.%d' % (col, cur_count)
counts[col] = cur_count + 1
else:
ncols = len(lines[0])
if not colNames:
columns = ['X.%d' % (i + 1) for i in range(ncols)]
else:
assert(len(colNames) == ncols)
columns = colNames
content = lines
if len(content) == 0: # pragma: no cover
if index_col is not None:
if np.isscalar(index_col):
index = Index([], name=columns.pop(index_col))
else:
cp_cols = list(columns)
names = []
for i in index_col:
name = cp_cols[i]
columns.remove(name)
names.append(name)
index = MultiIndex.fromarrays([[]] * len(index_col),
names=names)
else:
index = Index([])
return DataFrame(index=index, columns=columns)
# common NA values
# no longer excluding inf representations
# '1.#INF','-1.#INF', '1.#INF000000',
NA_VALUES = set(['-1.#IND', '1.#QNAN', '1.#IND', '-1.#QNAN',
'#N/A N/A', 'NA', '#NA', 'NULL', 'NaN',
'nan', ''])
if na_values is None:
na_values = NA_VALUES
else:
na_values = set(list(na_values)) | NA_VALUES
zipped_content = list(lib.to_object_array(content).T)
if index_col is None and len(content[0]) == len(columns) + 1:
index_col = 0
# no index column specified, so infer that's what is wanted
if index_col is not None:
if np.isscalar(index_col):
index = zipped_content.pop(index_col)
if len(content[0]) == len(columns) + 1:
name = None
else:
name = columns.pop(index_col)
else: # given a list of index
idx_names = []
index = []
for idx in index_col:
idx_names.append(columns[idx])
index.append(zipped_content[idx])
#remove index items from content and columns, don't pop in loop
for i in range(len(index_col)):
columns.remove(idx_names[i])
zipped_content.remove(index[i])
if np.isscalar(index_col):
if parse_dates:
index = lib.try_parse_dates(index, parser=date_parser)
index = Index(_convert_types(index, na_values), name=name)
else:
arrays = _maybe_convert_int_mindex(index, parse_dates,
date_parser)
index = MultiIndex.from_arrays(arrays,
names=idx_names)
else:
index = Index(np.arange(len(content)))
if not index._verify_integrity():
dups = index._get_duplicates()
raise Exception('Index has duplicates: %s' % str(dups))
if len(columns) != len(zipped_content):
raise Exception('wrong number of columns')
data = dict((k, v) for k, v in zip(columns, zipped_content))
data = _convert_to_ndarrays(data, na_values)
return DataFrame(data=data, columns=columns, index=index)
