Foreign table created in Parquet format usually contains the following columns:
- Location
- Several other user specified columns in Parquet format specified while creating foreign table
User can create a teradataml DataFrame on a foreign table using "DataFrame()" or "DataFrame.from_table()", the same way to create a teradataml DataFrame on a regular table. With the created DataFrame, user can easily access the data in these columns and process the data using teradataml DataFrame API or other Python packages.
How to access actual data and path variables
Unlike foreign tables on JSON and CSV format data, teradataml DataFrame on Parquet data provides direct access to the actual data in Parquet files, as described in Accessing Foreign Table Created On Parquet Data.
Path variables from the foreign table can be accessed in one of the following ways: