trasgodp.geoindis package

Module contents

Metric privacy mechanisms for location data.

trasgodp.geoindis.metric_privacy(df, column_lat, column_lon, epsilon, new_cols=False, earth_radius_m=6371000, seed=42)

Apply geo-indistinguishability to the columns of a DataFrame with lat, lon pairs.

Parameters:

• df (pd.DataFrame) – DataFrame containing the column to be disturbed.

• column_lat (str) – name of the column containing latitudes.

• column_lon (str) – name of the column containing longitudes.

• epsilon (float) – privacy budget.

• new_cols (bool) – if True, add new columns with the perturbed values. Otherwise, overwrite the original columns.

• earth_radius_m (float) – radius of the Earth in meters.

• seed (int) – random seed for reproducibility.

Returns:

DataFrame with the perturbed location data (including radius).

Return type:

pd.DataFrame

trasgodp.geoindis.plot_metric_dp_map(df_dp, column_lat, column_lon, save_file='metric_dp_map.html')

Plot (and save) a map with the original and perturbed locations, and the radius.

Parameters:

• df_dp (pd.DataFrame) – DataFrame containing both the original and perturbed data. This DataFrame must be generated including new_cols=True in the metric_privacy function.

• column_lon (str) – name of the column containing longitudes.

• column_lat (str) – name of the column containing latitudes.

• save_file (str) – name of the file where the map will be saved.

Returns:

map.

Return type:

folium.Map