PiArgus

This package provides a python wrapper around τ-ARGUS, a program to protect statistical tables. This package takes care of generating all the required metadata and runs τ-ARGUS in the background to do the heavy work.

For this package to work, it is required to install τ-ARGUS locally first. It's also recommended to read the TauArgus manual to understand how it should be used.

Features

• Generate output tables from microdata or tabledata. It is recommended to generate from microdata.
• Metadata can be generated automatically, although using an existing rda-file is also possible.
• It's possible to create hierarchies, codelists, apriori files, recode files all from code or from existing files.
• Basic error checking of input is done before input is supplied to argus.

Feel free to contribute for other TauArgus-features. Feedback is welcome too.

Installing

• Download and install the latest version of τ-ARGUS.
• Then use pip to install piargus:

$ pip install --upgrade piargus

Example

import pandas as pd
import piargus as pa

tau = pa.TauArgus(r'C:\Users\User\Programs\TauArgus4.2.0b5\TauArgus.exe')
input_df = pd.read_csv('data/microdata.csv')
input_data = pa.MicroData(input_df)
output_table = pa.Table(['sbi', 'regio'], 'income', safety_rule="P(10)")

job = pa.Job(input_data, [output_table], directory='tau')
report = tau.run(job)
table_result = output_table.load_result()

print(report)
print(table_result)

Change C:\Users\User\Programs\TauArgus4.2.0b5\TauArgus.exe to the location where argus is installed. See tutorial for a general introduction. See Examples for more examples.

See also

The following packages in R offer similar functionality:

• https://github.com/sdcTools/sdcTable
• https://github.com/InseeFrLab/rtauargus