def create_dataframe_multi_index(inp_path, section='CURVES'):
# format the section header for look up in headers OrderedDict
sect = remove_braces(section).upper()
# get list of all section headers in inp to use as section ending flags
headers = get_inp_sections_details(inp_path, include_brackets=False)
if sect not in headers:
warnings.warn(f'{sect} section not found in {inp_path}')
return pd.DataFrame()
# extract the string and read into a dataframe
start_string = format_inp_section_header(section)
end_strings = [format_inp_section_header(h) for h in headers.keys()]
s = extract_section_of_file(inp_path, start_string, end_strings)
cols = headers[sect]['columns']
f = StringIO(s)
data = []
for line in f.readlines():
items = line.strip().split()
if len(items) == 3:
items = [items[0], None, items[1], items[2]]
if len(items) == 4:
data.append(items)
df = pd.DataFrame(data=data, columns=cols)
df = df.set_index(['Name', 'Type'])
return df
def dataframe_from_inp(inp_path,
section,
additional_cols=None,
quote_replace=' ',
**kwargs):
"""
create a dataframe from a section of an INP file
:param inp_path:
:param section:
:param additional_cols:
:param skip_headers:
:param quote_replace:
:return:
"""
# format the section header for look up in headers OrderedDict
sect = remove_braces(section).upper()
# get list of all section headers in inp to use as section ending flags
headers = get_inp_sections_details(inp_path, include_brackets=False)
if sect not in headers:
warnings.warn(f'{sect} section not found in {inp_path}')
return pd.DataFrame()
# extract the string and read into a dataframe
start_string = format_inp_section_header(section)
end_strings = [format_inp_section_header(h) for h in headers.keys()]
s = extract_section_of_file(inp_path, start_string, end_strings, **kwargs)
# replace occurrences of double quotes ""
s = s.replace('""', quote_replace)
# and get the list of columns to use for parsing this section
# add any additional columns needed for special cases (build instructions)
additional_cols = [] if additional_cols is None else additional_cols
cols = headers[sect]['columns'] + additional_cols
if headers[sect]['columns'][0] == 'blob':
# return the whole row, without specific col headers
return pd.read_csv(StringIO(s), delim_whitespace=False)
else:
try:
df = pd.read_csv(StringIO(s),
header=None,
delim_whitespace=True,
skiprows=[0],
index_col=0,
names=cols)
except IndexError:
print(
f'failed to parse {section} with cols: {cols}. head:\n{s[:500]}'
)
raise
return df