snorkelflow.operators.pandas_operator

class snorkelflow.operators.pandas_operator(*, input_schema, name=None, resources=None, resources_fn=None, output_schema=None)

Bases: object

Decorator that wraps a function mapping a pandas.DataFrame to another pandas.DataFrame.

While @pandas_operator allows you to define custom operators using simple Pandas syntax, they are automatically executed and parallelized using Dask under the hood.

Examples

In the following example, a function that adds a random number to each row is defined. An output_schema is explicitly added to indicate the column schema of the resulting DataFrame.

This is used for validation and may provide you with helpful error messages.

from snorkelflow.operators import pandas_operator

@pandas_operator(name="Add Random Number", input_schema={}, output_schema={"rand_num": float})
def add_rand_num(df: pd.DataFrame) -> pd.DataFrame:
    import random
    df["rand_num"] = [random.random() for _ in range(len(df))]
    return df

sf.add_operator(add_rand_num)

Parameters:

• name (Optional[str], default: None) – Name of the Operator

• resources (Optional[Dict[str, Any]], default: None) – Resources passed in to f via kwargs

• resources_fn (Optional[Callable[[], Dict[str, Any]]], default: None) – A function for generating a dictionary of values passed to f via kwargs, that are too expensive to serialize as resources.

• input_schema (Dict[str, Any]) – Dictionary mapping from column to dtype, used to validate the dtypes of the input dataframe.

• output_schema (Optional[Dict[str, Any]], default: None) –

  Dictionary mapping from column to dtype, used to validate the dtypes of the output dataframe.

  If not None, then f must not delete any dataframe columns, and all new columns must be specified along with types in output_schema.

__init__(*, input_schema, name=None, resources=None, resources_fn=None, output_schema=None)

Methods

__init__(*, input_schema[, name, resources, ...])