def test_parse_feat_str_parse_valid_2(self):
feature_string = "jaccard(qgm_3(ltuple['zipcode']), qgm_3(ltuple['zipcode']))"
p_dict = _parse_feat_str(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
self.assertEqual(p_dict['left_attr_tokenizer'], 'qgm_3')
self.assertEqual(p_dict['right_attr_tokenizer'], 'qgm_3')
self.assertEqual(p_dict['simfunction'], 'jaccard')
def test_parse_feat_str_parse_exp(self):
feature_string = "jaccard~(qgm_3(ltuple[['zipcode']), qgm_3(rtuple['zipcode']))"
p_dict = _parse_feat_str(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
for k, v in six.iteritems(p_dict):
if k != 'is_auto_generated':
self.assertEqual(v, 'PARSE_EXP')
def test_add_feature_invalid_df_columns(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
add_feature(pd.DataFrame(), 'test', f_dict)
def test_parse_feat_str_parse_valid_1(self):
feature_string = "jaccard(qgm_3(ltuple['zipcode']), qgm_3(rtuple['zipcode']))"
p_dict = _parse_feat_str(feature_string, get_tokenizers_for_matching(), get_sim_funs_for_matching())
self.assertEqual(p_dict['left_attr_tokenizer'], 'qgm_3')
self.assertEqual(p_dict['right_attr_tokenizer'], 'qgm_3')
self.assertEqual(p_dict['simfunction'], 'jaccard')
self.assertEqual(p_dict['left_attribute'], 'zipcode')
self.assertEqual(p_dict['right_attribute'], 'zipcode')
def test_add_feature_name_already_present(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_table = create_feature_table()
len1 = len(feature_table)
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(), get_sim_funs_for_matching())
add_feature(feature_table, 'test', f_dict)
add_feature(feature_table, 'test', f_dict)
def test_get_features_invalid_ltable_rtable_switch(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
l_attr_types = au.get_attr_types(A)
r_attr_types = au.get_attr_types(B)
attr_corres = au.get_attr_corres(B, A)
tok = get_tokenizers_for_matching()
sim = get_sim_funs_for_matching()
feat_table = afg.get_features(A, B, l_attr_types, r_attr_types, attr_corres, tok, sim)
def test_add_feature_name_already_present(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_table = create_feature_table()
len1 = len(feature_table)
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
add_feature(feature_table, 'test', f_dict)
add_feature(feature_table, 'test', f_dict)
def test_get_features_invalid_ltable_rtable_switch(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
l_attr_types = au.get_attr_types(A)
r_attr_types = au.get_attr_types(B)
attr_corres = au.get_attr_corres(B, A)
tok = get_tokenizers_for_matching()
sim = get_sim_funs_for_matching()
feat_table = afg.get_features(A, B, l_attr_types, r_attr_types,
attr_corres, tok, sim)
def test_add_features_valid_1(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_table = get_features_for_matching(A, B, validate_inferred_attr_types=False)
len1 = len(feature_table)
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(), get_sim_funs_for_matching())
add_feature(feature_table, 'test', f_dict)
len2 = len(feature_table)
self.assertEqual(len1+1, len2)
self.assertEqual(feature_table.ix[len(feature_table)-1, 'function'](A.ix[1], B.ix[2]), 1.0)
def test_add_features_valid_1(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_table = get_features_for_matching(A, B)
len1 = len(feature_table)
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
add_feature(feature_table, 'test', f_dict)
len2 = len(feature_table)
self.assertEqual(len1 + 1, len2)
self.assertEqual(
feature_table.ix[len(feature_table) - 1, 'function'](A.ix[1],
B.ix[2]), 1.0)
def test_get_features_valid(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
l_attr_types = au.get_attr_types(A)
r_attr_types = au.get_attr_types(B)
attr_corres = au.get_attr_corres(A, B)
tok = get_tokenizers_for_matching()
sim = get_sim_funs_for_matching()
feat_table = afg.get_features(A, B, l_attr_types, r_attr_types, attr_corres, tok, sim)
self.assertEqual(isinstance(feat_table, pd.DataFrame), True)
functions = feat_table['function']
for f in functions:
x = f(A.ix[1], B.ix[2])
self.assertEqual(x >= 0, True)
def test_get_features_valid(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
l_attr_types = au.get_attr_types(A)
r_attr_types = au.get_attr_types(B)
attr_corres = au.get_attr_corres(A, B)
tok = get_tokenizers_for_matching()
sim = get_sim_funs_for_matching()
feat_table = afg.get_features(A, B, l_attr_types, r_attr_types,
attr_corres, tok, sim)
self.assertEqual(isinstance(feat_table, pd.DataFrame), True)
functions = feat_table['function']
for f in functions:
x = f(A.ix[1], B.ix[2])
self.assertEqual(x >= 0, True)
def test_add_feature_invalid_df_columns(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(), get_sim_funs_for_matching())
with self.assertRaises(AssertionError) as ctx:
add_feature(pd.DataFrame(), 'test', f_dict)
actual = str(ctx.exception)
print(actual)
expected = 'Feature table does not have all required columns\n ' \
'The following columns are missing: feature_name, left_attribute, right_attribute, ' \
'left_attr_tokenizer,' \
' right_attr_tokenizer, simfunction, function, function_source, is_auto_generated'
self.assertEqual(actual, expected)
def test_add_feature_invalid_df_columns(self):
A = read_csv_metadata(path_a)
B = read_csv_metadata(path_b, key='ID')
feature_string = "exact_match(ltuple['zipcode'], rtuple['zipcode'])"
f_dict = get_feature_fn(feature_string, get_tokenizers_for_matching(),
get_sim_funs_for_matching())
with self.assertRaises(AssertionError) as ctx:
add_feature(pd.DataFrame(), 'test', f_dict)
actual = str(ctx.exception)
print(actual)
expected = 'Feature table does not have all required columns\n ' \
'The following columns are missing: feature_name, left_attribute, right_attribute, ' \
'left_attr_tokenizer,' \
' right_attr_tokenizer, simfunction, function, function_source, is_auto_generated'
self.assertEqual(actual, expected)
def test_parse_feat_str_parse_exp(self):
feature_string = "jaccard~(qgm_3(ltuple[['zipcode']), qgm_3(rtuple['zipcode']))"
p_dict = _parse_feat_str(feature_string, get_tokenizers_for_matching(), get_sim_funs_for_matching())
for k,v in six.iteritems(p_dict):
if k != 'is_auto_generated':
self.assertEqual(v, 'PARSE_EXP')
def test_get_tokenizers_for_matching_invalid(self):
x = tok.get_tokenizers_for_matching(None, None)
def test_get_tokenizers_for_matching(self):
x = tok.get_tokenizers_for_matching()
self.assertEqual(isinstance(x, dict), True)
input = 'data science'
for name, value in six.iteritems(x):
self.assertEqual(isinstance(value(input), list), True)
def get_features_for_matching(ltable, rtable):
"""
This function automatically generates features that can be used for
matching purposes.
Args:
ltable,rtable (DataFrame): The pandas DataFrames for which the
features are to be generated.
Returns:
A pandas DataFrame containing automatically generated features.
Specifically, the DataFrame contains the following attributes:
'feature_name', 'left_attribute', 'right_attribute',
'left_attr_tokenizer', 'right_attr_tokenizer', 'simfunction',
'function', 'function_source', and 'is_auto_generated'.
Further, this function also sets the following global variables:
_match_t, _match_s, _atypes1, _atypes2, and _match_c.
The variable _match_t contains the tokenizers used and _match_s
contains the similarity functions used for creating features.
The variables _atypes1, and _atypes2 contain the attribute types for
ltable and rtable respectively. The variable _match_c contains the
attribute correspondences between the two input tables.
Raises:
AssertionError: If `ltable` is not of type pandas
DataFrame.
AssertionError: If `rtable` is not of type pandas
DataFrame.
Note:
In the output DataFrame, two
attributes demand some explanation: (1) function, and (2)
is_auto_generated. The function, points to the actual Python function
that implements the feature. Specifically, the function takes in two
tuples (one from each input table) and returns a numeric value. The
attribute is_auto_generated contains either True or False. The flag
is True only if the feature is automatically generated by py_entitymatching.
This is important because this flag is used to make some assumptions
about the semantics of the similarity function used and use that
information for scaling purposes.
See Also:
:meth:`py_entitymatching.get_attr_corres`, :meth:`py_entitymatching.get_attr_types`,
:meth:`py_entitymatching.get_sim_funs_for_matching`
:meth:`py_entitymatching.get_tokenizers_for_matching`
"""
# Validate input parameters
# # We expect the ltable to be of type pandas DataFrame
if not isinstance(ltable, pd.DataFrame):
logger.error('Input table A is not of type pandas DataFrame')
raise AssertionError('Input table A is not of type pandas DataFrame')
# # We expect the rtable to be of type pandas DataFrame
if not isinstance(rtable, pd.DataFrame):
logger.error('Input table B is not of type pandas DataFrame')
raise AssertionError('Input table B is not of type pandas DataFrame')
# Get similarity functions for generating the features for matching
sim_funcs = sim.get_sim_funs_for_matching()
# Get tokenizer functions for generating the features for matching
tok_funcs = tok.get_tokenizers_for_matching()
# Get the attribute types of the input tables
attr_types_ltable = au.get_attr_types(ltable)
attr_types_rtable = au.get_attr_types(rtable)
# Get the attribute correspondence between the input tables
attr_corres = au.get_attr_corres(ltable, rtable)
# Get the features
feature_table = get_features(ltable, rtable, attr_types_ltable,
attr_types_rtable, attr_corres, tok_funcs,
sim_funcs)
# Export important variables to global name space
em._match_t = tok_funcs
em._match_s = sim_funcs
em._atypes1 = attr_types_ltable
em._atypes2 = attr_types_ltable
em._match_c = attr_corres
# Finally return the feature table
return feature_table
def get_features_for_matching(ltable, rtable, validate_inferred_attr_types=True):
"""
This function automatically generates features that can be used for
matching purposes.
Args:
ltable,rtable (DataFrame): The pandas DataFrames for which the
features are to be generated.
validate_inferred_attr_types (boolean): A flag to indicate whether to
show the user the inferred attribute types and the features
chosen for those types.
Returns:
A pandas DataFrame containing automatically generated features.
Specifically, the DataFrame contains the following attributes:
'feature_name', 'left_attribute', 'right_attribute',
'left_attr_tokenizer', 'right_attr_tokenizer', 'simfunction',
'function', 'function_source', and 'is_auto_generated'.
Further, this function also sets the following global variables:
_match_t, _match_s, _atypes1, _atypes2, and _match_c.
The variable _match_t contains the tokenizers used and _match_s
contains the similarity functions used for creating features.
The variables _atypes1, and _atypes2 contain the attribute types for
ltable and rtable respectively. The variable _match_c contains the
attribute correspondences between the two input tables.
Raises:
AssertionError: If `ltable` is not of type pandas
DataFrame.
AssertionError: If `rtable` is not of type pandas
DataFrame.
AssertionError: If `validate_inferred_attr_types` is not of type
pandas DataFrame.
Examples:
>>> import py_entitymatching as em
>>> A = em.read_csv_metadata('path_to_csv_dir/table_A.csv', key='ID')
>>> B = em.read_csv_metadata('path_to_csv_dir/table_B.csv', key='ID')
>>> match_f = em.get_features_for_matching(A, B)
Note:
In the output DataFrame, two
attributes demand some explanation: (1) function, and (2)
is_auto_generated. The function, points to the actual Python function
that implements the feature. Specifically, the function takes in two
tuples (one from each input table) and returns a numeric value. The
attribute is_auto_generated contains either True or False. The flag
is True only if the feature is automatically generated by py_entitymatching.
This is important because this flag is used to make some assumptions
about the semantics of the similarity function used and use that
information for scaling purposes.
See Also:
:meth:`py_entitymatching.get_attr_corres`, :meth:`py_entitymatching.get_attr_types`,
:meth:`py_entitymatching.get_sim_funs_for_matching`
:meth:`py_entitymatching.get_tokenizers_for_matching`
"""
# Validate input parameters
# # We expect the ltable to be of type pandas DataFrame
validate_object_type(ltable, pd.DataFrame, 'Input table A')
# # We expect the rtable to be of type pandas DataFrame
validate_object_type(rtable, pd.DataFrame, 'Input table B')
# # We expect the validate_inferred_attr_types to be of type boolean
validate_object_type(validate_inferred_attr_types, bool, 'Validate inferred attribute type')
# Get similarity functions for generating the features for matching
sim_funcs = sim.get_sim_funs_for_matching()
# Get tokenizer functions for generating the features for matching
tok_funcs = tok.get_tokenizers_for_matching()
# Get the attribute types of the input tables
attr_types_ltable = au.get_attr_types(ltable)
attr_types_rtable = au.get_attr_types(rtable)
# Get the attribute correspondence between the input tables
attr_corres = au.get_attr_corres(ltable, rtable)
# Show the user inferred attribute types and features and request
# user permission to proceed
if validate_inferred_attr_types:
# if the user does not want to proceed, then exit the function
if validate_attr_types(attr_types_ltable, attr_types_rtable, attr_corres) is None:
return
# Get the features
feature_table = get_features(ltable, rtable, attr_types_ltable,
attr_types_rtable, attr_corres,
tok_funcs, sim_funcs)
# Export important variables to global name space
em._match_t = tok_funcs
em._match_s = sim_funcs
em._atypes1 = attr_types_ltable
em._atypes2 = attr_types_rtable
em._match_c = attr_corres
# Finally return the feature table
return feature_table
