def check_for_new_cas(tab_manager=None):
# Make a list of CAS numbers to run through SciFinder
n_p = pd.DataFrame({'not_perf':cat_rec.Categorize_CAS(tab_manager=tab_manager).get_corrected_not_perf()})
n_p['keep'] = n_p.not_perf.map(lambda x : ct.is_valid_CAS_code(x))
n_p[n_p.keep].to_csv(tmpdir+'cas_to_check.csv',index=False)
def _clean_CAS_for_comparison(self):
#print('clean cas for comparison')
#self.cas_field_cat.rename({'original':'CASNumber'},inplace=True,axis=1)
self.cas_field_cat[
'cas_clean'] = self.cas_field_cat.CASNumber.str.replace(
r'[^0-9-]', '')
self.cas_field_cat[
'zero_corrected'] = self.cas_field_cat.cas_clean.map(
lambda x: ct.correct_zeros(x))
def examine_not_valid(self):
# !!!! DONT USE THIS YET
cntr = 1
for tup in self.reviewed:
if self.reviewed[tup] == False:
if not ct.is_valid_CAS_code(tup[0]):
print('\n\n', tup, cntr, '\n\n')
cntr += 1
self._add_to_results_dic(tup[0], tup[1], 'non_valid_cas',
8)
def _get_field_dic(self):
"""takes the precompiled field_dic and creates casig_orig df from it."""
with open(field_dic_fn, 'rb') as f:
field_dic = pickle.load(f)
self.casig_orig = field_dic['casig']
self.casig_orig['cas_clean'] = self.casig_orig.CASNumber.str.replace(
r'[^0-9-]', '')
self.casig_orig['zero_corrected'] = self.casig_orig.cas_clean.map(
lambda x: ct.correct_zeros(x))
# now the Ingredients
self.casig_orig['ig_clean'] = self.casig_orig.IngredientName.str.strip(
).str.lower()
self.casig_orig.drop_duplicates(inplace=True)
self.casig_orig.to_csv(outdir + 'temp.csv',
quotechar='$',
quoting=csv.QUOTE_ALL)
def _clean_CAS_for_comparison(self):
self.cas_field_cat['cas_clean'] = self.cas_field_cat.CASNumber.str.replace(r'[^0-9-]','')
self.cas_field_cat['zero_corrected'] = self.cas_field_cat.cas_clean.map(lambda x: ct.correct_zeros(x) )
# replace the handful of obsolete cas numbers with widely used number
self.cas_field_cat.zero_corrected = self.cas_field_cat.zero_corrected.map(lambda x: self._replace_obsolete(x))