auto_code#

tdprepview.auto_code #

auto_code(DF, input_schema='', input_table='', non_feature_cols=[])

Generate Python code for a suggested preprocessing pipeline based on a DataFrame or database schema.

This function analyzes the input data or the database schema to automatically create a tdprepview preprocessing pipeline. The output is Python code that can serve as a starting point for building machine learning workflows in ClearScape.

Parameters:

Name	Type	Description	Default
`DF`	`DataFrame`	A `teradataml.DataFrame` containing the input data. If `None`, the DataFrame will be constructed from `input_schema` and `input_table`.	required
`input_schema`	`str`	Database schema name, used to generate the DataFrame if `DF` is `None`.	`''`
`input_table`	`str`	Database table or view name, used to generate the DataFrame if `DF` is `None`.	`''`
`non_feature_cols`	`list`	List of column names to exclude from preprocessing, e.g., IDs or target variables.	`[]`

Returns:

Type	Description
`str`	A string containing Python code that defines the suggested preprocessing pipeline.

Examples:

Generate code from a DataFrame:

import tdprepview
import teradataml as tdml

DF = tdml.DataFrame(tdml.in_schema("my_schema","my_table"))
code_str = tdprepview.auto_code(DF, non_feature_cols=["ID","target"])
print(code_str)
my_pipeline = tdprepview.Pipeline(steps= eval(code_str))

Generate code from database table directly:

import tdprepview

code_str = tdprepview.auto_code(
    DF=None,
    input_schema="my_schema",
    input_table="my_table",
    non_feature_cols=["ID","target"]
)
print(code_str)
my_pipeline = tdprepview.Pipeline(steps= eval(code_str))