The package generates semantically and syntactically correct random SELECT SQL Statements. It is developed by EHIL mainly to check the query response from synthetic data against that of real data. The package supports both tabular and longitudinal datasets. Table shapes, variable names and relations in both real and synthetic datasets shall be identical.
To install:
pip install fuzzy-sql
For further details, please refer to the Documentation.
Detailed examples comprising three sample datasets are provided under the examples folder in the repository. To generate the random queries, you first need to download the sample data by running 0.0-download_data.ipynb. You may then proceed with the remaining notebooks to construct the necessary databases and generate the random queries. For details, please refer to the usage/code-examples subsection in the Documentation.