Jaa


Convert between PySpark and pandas DataFrames

Learn how to convert Apache Spark DataFrames to and from pandas DataFrames using Apache Arrow in Azure Databricks.

Apache Arrow and PyArrow

Apache Arrow is an in-memory columnar data format used in Apache Spark to efficiently transfer data between JVM and Python processes. This is beneficial to Python developers who work with pandas and NumPy data. However, its usage requires some minor configuration or code changes to ensure compatibility and gain the most benefit.

PyArrow is a Python binding for Apache Arrow and is installed in Databricks Runtime. For information on the version of PyArrow available in each Databricks Runtime version, see the Databricks Runtime release notes versions and compatibility.

Supported SQL types

All Spark SQL data types are supported by Arrow-based conversion except ArrayType of TimestampType. MapType and ArrayType of nested StructType are only supported when using PyArrow 2.0.0 and above. StructType is represented as a pandas.DataFrame instead of pandas.Series.

Convert PySpark DataFrames to and from pandas DataFrames

Arrow is available as an optimization when converting a PySpark DataFrame to a pandas DataFrame with toPandas() and when creating a PySpark DataFrame from a pandas DataFrame with createDataFrame(pandas_df).

To use Arrow for these methods, set the Spark configuration spark.sql.execution.arrow.pyspark.enabled to true. This configuration is enabled by default except for High Concurrency clusters as well as user isolation clusters in workspaces that are Unity Catalog enabled.

In addition, optimizations enabled by spark.sql.execution.arrow.pyspark.enabled could fall back to a non-Arrow implementation if an error occurs before the computation within Spark. You can control this behavior using the Spark configuration spark.sql.execution.arrow.pyspark.fallback.enabled.

Example

import numpy as np
import pandas as pd

# Enable Arrow-based columnar data transfers
spark.conf.set("spark.sql.execution.arrow.pyspark.enabled", "true")

# Generate a pandas DataFrame
pdf = pd.DataFrame(np.random.rand(100, 3))

# Create a Spark DataFrame from a pandas DataFrame using Arrow
df = spark.createDataFrame(pdf)

# Convert the Spark DataFrame back to a pandas DataFrame using Arrow
result_pdf = df.select("*").toPandas()

Using the Arrow optimizations produces the same results as when Arrow is not enabled. Even with Arrow, toPandas() results in the collection of all records in the DataFrame to the driver program and should be done on a small subset of the data.

In addition, not all Spark data types are supported and an error can be raised if a column has an unsupported type. If an error occurs during createDataFrame(), Spark creates the DataFrame without Arrow.