Implementation:Eventual Inc Daft Read Deltalake

Knowledge Sources	Daft Daft Docs
Domains	Data_Engineering, Data_Lakehouse
Last Updated	2026-02-08 00:00 GMT

Overview

Concrete tool for reading Delta Lake tables into a lazy distributed DataFrame provided by the Daft library.

Description

The read_deltalake function creates a lazy DataFrame scan of a Delta Lake table. It accepts table URIs (including remote object stores like s3:// and gs://), DataCatalogTable references (e.g., AWS Glue), and UnityCatalogTable instances. Time travel is supported via version numbers (int), RFC 3339 timestamps (string), or datetime objects. The function creates a DeltaLakeScanOperator that handles the transaction log parsing, deletion vector processing, and predicate pushdown. Multithreaded IO is automatically disabled when running on the Ray runner to reduce system resource contention.

Usage

Import and use this function when you need to read data from a Delta Lake table with versioned access and deletion vector support.

Code Reference

Source Location

Repository: Daft
File: daft/io/delta_lake/_deltalake.py
Lines: L24-98

Signature

def read_deltalake(
    table: Union[str, DataCatalogTable, "UnityCatalogTable"],
    version: Union[int, str, "datetime"] | None = None,
    io_config: IOConfig | None = None,
    ignore_deletion_vectors: bool = False,
    _multithreaded_io: bool | None = None,
) -> DataFrame

Import

from daft import read_deltalake

# or
import daft
df = daft.read_deltalake(uri)

I/O Contract

Inputs

Name	Type	Required	Description
table	DataCatalogTable \| UnityCatalogTable	Yes	URI to Delta Lake table or a catalog table reference
version	str \| datetime \| None	No	Version number, RFC 3339 timestamp string, or datetime for time travel; defaults to latest version
io_config	None	No	Custom IO configuration for accessing object storage; defaults to Daft context config
ignore_deletion_vectors	bool	No	Whether to skip checking deletion vectors when reading; defaults to False

Outputs

Name	Type	Description
return	DataFrame	A lazy DataFrame with the schema converted from the Delta Lake table

Usage Examples

Basic Usage

import daft

# Read a Delta Lake table from a local path
df = daft.read_deltalake("some-table-uri")

# Apply filters (pushed down through Delta Lake metadata)
df = df.where(df["foo"] > 5)
df.show()

# Read from S3 with custom IO config
from daft.io import S3Config, IOConfig
io_config = IOConfig(s3=S3Config(region="us-west-2", anonymous=True))
df = daft.read_deltalake("s3://daft-public-data/test_fixtures/delta_table/", io_config=io_config)

# Read a specific version for time travel
df = daft.read_deltalake("s3://bucket/table", version=5)

Related Pages

Implements Principle

Principle:Eventual_Inc_Daft_Delta_Lake_Reading

Requires Environment

Uses Heuristic

Heuristic:Eventual_Inc_Daft_Delta_Lake_S3_Locking

Page Connections

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

Principle

Implementation

Heuristic

Environment