Tabular data is often transformed in a flattened manner so it can easily be processed by commen ML frameworks. Sometimes it is usefull to understand the relationships and hierarchies between the columns of a flattend datasource. In this notebook hierarchies are extracted from a flattened csv file.