nvtabular.ops.LambdaOp

class nvtabular.ops.LambdaOp(f, dependency=None, label=None, dtype=None, tags=None, properties=None)[source]

Bases: nvtabular.ops.operator.Operator

LambdaOp allows you to apply row level functions to an NVTabular workflow.

Example usage 1:

# Define a ColumnSelector that LamdaOp will apply to
# then define a custom function, e.g. extract first 5 character from a string
lambda_feature = ColumnSelector(["col1"])
new_lambda_feature = lambda_feature >> (lambda col: col.str.slice(0, 5))
processor = nvtabular.Workflow(new_lambda_feature + 'label')

Example usage 2:

# define a custom function e.g. calculate probability for different events.
# Rename the each new feature column name.
lambda_features = ColumnSelector(['event1', 'event2', 'event3']), # columns, f is applied to
def cond_prob(col, gdf):
    col = col.astype(np.float32)
    col = col / gdf['total_events']
    return col
new_lambda_features = lambda_features >> LambdaOp(cond_prob, dependency=["total_events"]) >> Rename(postfix="_cond")
processor = nvtabular.Workflow(new_lambda_features + 'label')

Parameters

f (callable) – Defines a function that takes a Series and an optional DataFrame as input, and returns a new Series as the output.
dependency (list, default None) – Whether to provide a dependency column or not.

__init__(f, dependency=None, label=None, dtype=None, tags=None, properties=None)[source]

Methods

`__init__`(f[, dependency, label, dtype, …])
`column_mapping`(col_selector)
`compute_column_schema`(col_name, input_schema)
`compute_input_schema`(root_schema, …)	Given the schemas coming from upstream sources and a column selector for the input columns, returns a set of schemas for the input columns this operator will use :param root_schema: Base schema of the dataset before running any operators.
`compute_output_schema`(input_schema, col_selector)	Given a set of schemas and a column selector for the input columns, returns a set of schemas for the transformed columns this operator will produce :param input_schema: The schemas of the columns to apply this operator to :type input_schema: Schema :param col_selector: The column selector to apply to the input schema :type col_selector: ColumnSelector
`compute_selector`(input_schema, selector, …)
`create_node`(selector)
`inference_initialize`(col_selector, model_config)	Configures this operator for use in inference.
`output_column_names`(col_selector)	Given a set of columns names returns the names of the transformed columns this operator will produce :param columns: The columns to apply this operator to :type columns: list of str, or list of list of str
`transform`(col_selector, df)	Transform the dataframe by applying this operator to the set of input columns

Attributes

`dependencies`
`dynamic_dtypes`
`label`
`output_dtype`
`output_properties`
`output_tags`
`supports`	Returns what kind of data representation this operator supports

transform(col_selector: merlin.dag.selector.ColumnSelector, df: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame[source]

Transform the dataframe by applying this operator to the set of input columns

Parameters

columns (list of str or list of list of str) – The columns to apply this operator to
df (Dataframe) – A pandas or cudf dataframe that this operator will work on

Returns

Returns a transformed dataframe for this operator

Return type

DataFrame

property dependencies

property label

column_mapping(col_selector)[source]

property dynamic_dtypes

property output_dtype

property output_tags

property output_properties