Implementation:Evidentlyai Evidently TextEvals Preset

Knowledge Sources	Evidently Evidently Text Evals
Domains	NLP, Text_Analysis
Last Updated	2026-02-14 12:00 GMT

Overview

Concrete preset container for summarizing text descriptor statistics provided by the Evidently library.

Description

TextEvals is a MetricContainer that generates value statistics for all text descriptor columns. It expands into ValueStats metrics for each numerical and categorical descriptor column, plus RowCount and RowTestSummary.

Usage

Include in a Report after creating a dataset with text descriptors.

Code Reference

Source Location

• Repository: evidently
• File: src/evidently/presets/dataset_stats.py
• Lines: L494-550

Signature

class TextEvals(MetricContainer):
    def __init__(
        self,
        columns: Optional[List[str]] = None,
        row_count_tests: GenericSingleValueMetricTests = None,
        column_tests: Optional[Dict[str, ValueStatsTests]] = None,
        include_tests: bool = True,
    ):
        """
        Args:
            columns: Optional list of descriptor column names to analyze (None = all).
            row_count_tests: Optional test conditions for row count.
            column_tests: Optional dict of column name to test configurations.
            include_tests: Whether to include auto-tests (default: True).
        """

Import

from evidently.presets import TextEvals

I/O Contract

Inputs

Name	Type	Required	Description
columns	Optional[List[str]]	No	Descriptor columns to analyze (None = all descriptors)
row_count_tests	GenericSingleValueMetricTests	No	Test conditions for row count
column_tests	Optional[Dict[str, ValueStatsTests]]	No	Per-column test configs
include_tests	bool	No	Enable auto-tests (default: True)

Outputs

Name	Type	Description
RowCount	int	Number of rows in the dataset
ValueStats per column	dict	Mean, std, min, max, quantiles for each descriptor column
RowTestSummary	dict	Summary of row-level test results

Usage Examples

from evidently import Report, Dataset, DataDefinition
from evidently.descriptors import TextLength, Sentiment, OOVWordsPercentage
from evidently.presets import TextEvals

# 1. Create dataset with descriptors
dataset = Dataset.from_pandas(
    df,
    data_definition=DataDefinition(text_columns=["Review_Text"]),
    descriptors=[
        TextLength("Review_Text"),
        Sentiment("Review_Text"),
        OOVWordsPercentage("Review_Text"),
    ],
)

# 2. Build report with TextEvals
report = Report([TextEvals()])
snapshot = report.run(dataset, reference_dataset)
snapshot.save_html("text_quality_report.html")

Related Pages

Implements Principle

• Principle:Evidentlyai_Evidently_Text_Quality_Reporting

Requires Environment

• Environment:Evidentlyai_Evidently_LLM_Evaluation_Environment