def __ncp_set_valued(original_series, anonymized_series):
original_flattened, original_indexes, _ = flatten_set_valued_series(
original_series)
anonymized_flattened, anonymized_indexes, _ = flatten_set_valued_series(
anonymized_series)
if is_categorical_dtype(original_series):
original_flattened_series = pd.Series(original_flattened,
index=original_indexes,
dtype="category",
name=original_series.name)
else:
original_flattened_series = pd.Series(original_flattened,
index=original_indexes,
name=original_series.name)
if is_categorical_dtype(anonymized_series):
anonymized_flattened_series = pd.Series(anonymized_flattened,
index=anonymized_indexes,
dtype="category",
name=anonymized_series.name)
else:
anonymized_flattened_series = pd.Series(anonymized_flattened,
index=anonymized_indexes,
name=anonymized_series.name)
return __calculate_ncp_attribute(original_flattened_series,
anonymized_flattened_series)
def recode_set_valued(series, recoding_rules, hierarchies):
"""
Generalizes set valued series by flattening
Parameters
----------
series: Series
Series to be recoded.
recoding_rules: dict
Dictionary containing recoding rules.
hierarchies: dict
Dictionary containing generalization hierarchies.
Returns
-------
any
Single value recoded to.
"""
flattened, indexes, is_category = flatten_set_valued_series(series)
if is_categorical_dtype(series) or is_category:
flattened_series = pd.Series(flattened,
index=indexes,
dtype="category",
name=series.name)
else:
flattened_series = pd.Series(flattened,
index=indexes,
name=series.name)
result = recode(flattened_series, recoding_rules, hierarchies)
return result.iloc[0]
def __calculate_ncp_attribute(original_series, anonymized_series):
if must_be_flattened(original_series):
original_flattened, original_indexes, is_category = flatten_set_valued_series(
original_series)
if is_categorical_dtype(original_series) or is_category:
original_flattened_series = pd.Series(original_flattened,
index=original_indexes,
dtype="category",
name=original_series.name)
else:
original_flattened_series = pd.Series(original_flattened,
index=original_indexes,
name=original_series.name)
ncp = __calculate_ncp_attribute(original_flattened_series,
anonymized_series)
elif is_node(anonymized_series): # Has been anonymized using a hierarchy
ncp = __ncp_numerical_hierarchy(original_series, anonymized_series)
elif is_datetime64_any_dtype(original_series):
ncp = __ncp_date(original_series, anonymized_series)
elif is_categorical_dtype(original_series):
ncp = __ncp_categorical(original_series, anonymized_series)
elif is_numeric_dtype(original_series):
ncp = __ncp_numerical(original_series, anonymized_series)
elif is_token_list(original_series):
ncp = __ncp_tokens(original_series, anonymized_series)
else:
ncp = __ncp_set_valued(original_series, anonymized_series)
return ncp
def __get_attribute_span(series):
if is_categorical_dtype(series):
span = len(series.unique())
elif is_datetime64_any_dtype(series):
span = series.max() - series.min()
span = span.days
elif is_numeric_dtype(series):
span = series.max() - series.min()
else:
flattened, indexes, is_category = flatten_set_valued_series(series)
if is_category:
new_series = pd.Series(flattened, dtype="category", index=indexes, name=series.name)
new_series.index.name = "id"
grouped = new_series.groupby(by="id").agg(agg_categorical).astype('category')
else:
new_series = pd.Series(flattened, index=indexes, name=series.name)
new_series.index.name = "id"
grouped = new_series.groupby(by="id").agg(agg_mean)
span = __get_attribute_span(grouped)
return span
def __split_partition(series):
if is_categorical_dtype(series) or is_datetime64_any_dtype(series):
values = series.sort_values().unique()
lv = set(values[:len(values) // 2])
rv = set(values[len(values) // 2:])
return series.index[series.isin(lv)], series.index[series.isin(rv)]
elif is_numeric_dtype(series):
median = series.median()
dfl = series.index[series < median]
dfr = series.index[series >= median]
return (dfl, dfr)
else:
flattened, indexes, is_category = flatten_set_valued_series(series)
if is_category:
new_series = pd.Series(flattened, index=indexes, dtype="category", name=series.name)
new_series.index.name = "id"
grouped = new_series.groupby(by="id").agg(agg_categorical).astype('category')
else:
new_series = pd.Series(flattened, index=indexes, name=series.name)
new_series.index.name = "id"
grouped = new_series.groupby(by="id").agg(agg_mean)
return __split_partition(grouped)