def write_filter_file(self, df):
filter_file_split = self.filter_file.split('::')
filter_file_name = filter_file_split[0]
if len(filter_file_split) > 1:
filter_file_cols = filter_file_split[1].split(',')
else:
filter_file_cols = []
df = df[[self.ad_col, self.ad_group_col]]
ndf = df[[self.ad_col]].drop_duplicates().reset_index(drop=True)
for col in filter_file_cols:
col = int(col)
tdf = pd.DataFrame(df[self.ad_group_col].str.split('_').str[col])
new_col_name = 'adset_name::{}'.format(col)
tdf = tdf.rename(columns={self.ad_group_col: new_col_name})
tdf = tdf.join(df[[self.ad_col
]]).drop_duplicates().reset_index(drop=True)
fdf = pd.DataFrame()
for value in tdf[self.ad_col].unique():
unique_vals = '|'.join(tdf[tdf[self.ad_col] == value]
[new_col_name].unique().tolist())
value_dict = {
new_col_name: [unique_vals],
self.ad_col: [value]
}
fdf = fdf.append(pd.DataFrame(value_dict),
ignore_index=True,
sort=False)
ndf = ndf.merge(fdf, on=self.ad_col)
ndf = ndf.rename(columns={self.ad_col: 'ad_name'})
utl.write_df(ndf, file_name=utl.config_file_path + filter_file_name)
def append_and_write_relation_df(self, relation_df):
df = pd.read_excel(utl.config_file_path + self.relation_file,
dtype=object,
keep_default_na=False,
na_values=[''])
df = df[~df[Creator.rel_col_imp].
isin(['creative_filename', 'body', 'description', 'title'])]
df = df.append(relation_df, ignore_index=True, sort=False)
df = df.drop_duplicates()
utl.write_df(df, utl.config_file_path + self.relation_file)
def get_plan_names(self):
self.col_name = self.col_name.split('|')
df_dict = {}
for col in self.col_name:
if col in self.df.columns:
df_dict[col] = pd.Series(self.df[col].unique())
else:
logging.warning('{} not in df. Continuing.'.format(col))
ndf = pd.DataFrame(df_dict)
utl.write_df(ndf, './' + self.new_file)
def apply_duplication(self):
cdf = pd.read_excel(self.new_file)
original_cols = cdf.columns
duplicated_col = self.col_name.split('::')[0]
unique_list = cdf[duplicated_col].unique()
cdf = pd.DataFrame(columns=original_cols)
self.df = self.df[self.col_name.split('::')[1].split('|')][:]
for item in unique_list:
self.df[duplicated_col] = item
cdf = cdf.append(self.df, ignore_index=True, sort=False)
cdf = cdf.reset_index(drop=True)
cdf = cdf[original_cols]
if len(self.col_name.split('::')) > 2:
cdf = self.apply_upload_filter(cdf)
utl.write_df(cdf, self.new_file)
def apply_relations(self):
cdf = pd.read_excel(self.new_file)
for imp_col in self.df['impacted_column_name'].unique():
df = self.df[self.df['impacted_column_name'] == imp_col]
par_col = str(df['column_name'].values[0]).split('|')
position = str(df['position'].values[0]).split('|')
if position == ['Constant']:
cdf[imp_col] = df['impacted_column_new_value'].values[0]
else:
rel_dict = self.create_relation_dictionary(df)
cdf = self.set_values_to_imp_col(cdf, position, par_col,
imp_col)
cdf = self.check_undefined_relation(cdf, rel_dict, imp_col)
cdf[imp_col] = cdf[imp_col].replace(rel_dict)
utl.write_df(cdf, self.new_file)
return self.error_dict
def check_undefined_relation(self, df, rel_dict, imp_col):
undefined = df.loc[~df[imp_col].isin(rel_dict), imp_col]
imp_file = self.new_file.split('.')[0].replace(file_path, '')
file_name = utl.err_file_path + imp_file + '_' + imp_col + '.xlsx'
if not undefined.empty:
logging.warning('No match found for the following values, '
'they were left blank. An error report was '
'generated ' + str(undefined.head().values))
df.loc[~df[imp_col].isin(rel_dict), imp_col] = ''
self.error_dict[imp_col] = len(undefined.unique())
err_file_path = os.path.join(
*[x for x in file_name.split('/') if '.' not in x])
utl.dir_check(err_file_path)
utl.write_df(undefined.drop_duplicates(), file_name)
else:
self.error_dict[imp_col] = 0
utl.remove_file(file_name)
return df
def write_name_creator_file(self):
df = pd.DataFrame(self.df[self.ad_col].unique(), columns=[self.ad_col])
utl.write_df(df, file_name=utl.config_file_path + self.creator_file)
def create_upload_file(self):
combined_list = self.get_combined_list()
df = self.create_df(combined_list)
utl.write_df(df, self.new_file)