Getting started

This first example uses the adult dataset. The idea is to apply local DP, first, to a numerical attribute (age) and second, to a categorical one (workclass). The resulting values will be stored in two new columns of the dataframe.

import pandas as pd
from trasgodp.numerical import dp_clip_laplace
from trasgodp.categorical import dp_exponential

# Read and process the data
data = pd.read_csv("examples/adult.csv")
data.columns = data.columns.str.strip()
cols = [
    "workclass",
    "education",
    "marital-status",
    "occupation",
    "sex",
    "native-country",
]
for col in cols:
    data[col] = data[col].str.strip()

# Apply DP for the attribute age:
column_num = "age"
epsilon1 = 10
df_age = dp_clip_laplace(data, column_num, epsilon1, new_column=True)

# Apply DP for the attribute workclass:
column_cat = "workclass"
epsilon2 = 5
df = dp_exponential(df_age, column_cat, epsilon2, new_column=True)

The second example uses the earthquake dataset for showcasing the use of metric privacy in location-based data (with latitude and longite values for each row). The geo-indistinguishability approach is applied and the map with original and privatized values is generated:

import pandas as pd
from trasgodp.geoindis import metric_privacy, plot_metric_dp_map

# Read the data
data = pd.read_csv("./examples/earthquake_data.csv")
column_lat = "latitude"
column_lon = "longitude"

# Apply metric privacy creating new columns for lat and lon:
epsilon =1.e-3
data_priv = metric_privacy(data, column_lat, column_lon, epsilon, new_cols=True)

# Plot and save the map:
plot_metric_dp_map(data_priv, column_lat, column_lon, save_file="example_map.html")