Implementation:Evidentlyai Evidently Text Length Descriptor

Knowledge Sources	Evidentlyai_Evidently
Domains	Descriptors, Text Analysis, Data Monitoring
Last Updated	2026-02-14 12:00 GMT

Overview

Descriptor that computes the character length of text values in each row of a specified column, returning a numerical column suitable for monitoring and testing.

Description

The TextLength class is a Descriptor subclass that measures the length of text in a given column. For each row, it applies a helper function _apply that returns:

0 for None or NaN values
len(value) for all other values

The result is a DatasetColumn of type ColumnType.Numerical, making it compatible with numerical metrics and tests.

The implementation converts the dataset to a DataFrame, applies the _apply function row-wise using pandas.Series.apply, and wraps the result in a DatasetColumn.

Usage

Use TextLength when you need to monitor or test the length of text content in your data. Common use cases include checking for empty responses, monitoring prompt/response lengths in LLM applications, or detecting anomalous text lengths in production data.

Code Reference

Source Location

Repository: Evidentlyai_Evidently
File: src/evidently/descriptors/_text_length.py

Signature

class TextLength(Descriptor):
    column_name: str

    def __init__(
        self, column_name: str,
        alias: Optional[str] = None,
        tests: Optional[List[AnyDescriptorTest]] = None,
    ):
    def generate_data(self, dataset: "Dataset", options: Options) -> Union[DatasetColumn, Dict[str, DatasetColumn]]:
    def list_input_columns(self) -> Optional[List[str]]:

def _apply(value: Any):

Import

from evidently.descriptors._text_length import TextLength

I/O Contract

Inputs

Name	Type	Required	Description
column_name	str	Yes	Name of the text column to measure length for
alias	Optional[str]	No (default: "text_length")	Custom display name for the descriptor
tests	Optional[List[AnyDescriptorTest]]	No	Tests to apply to the computed text length values

Outputs

Name	Type	Description
return	DatasetColumn	Numerical column containing character counts per row; 0 for None/NaN values

Usage Examples

from evidently.descriptors._text_length import TextLength

# Basic usage - measure length of a text column
descriptor = TextLength(column_name="response")

# With custom alias and tests
from evidently.core.datasets import gt
descriptor = TextLength(
    column_name="response",
    alias="response_length",
    tests=[gt(0)],  # ensure all responses are non-empty
)

# Add to dataset
dataset.add_descriptors([descriptor])

# Use in a report
from evidently import Report
from evidently.metrics import MeanValue
report = Report(metrics=[MeanValue("response_length")])
snapshot = report.run(dataset)

Related Pages

Environment:Evidentlyai_Evidently_Python_Core_Environment
Evidentlyai_Evidently_Custom_Descriptors - Custom descriptor classes for user-defined transformations
Evidentlyai_Evidently_Context_Relevance - Context relevance descriptor for RAG evaluation

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment