/koalas

Koalas: pandas API on Apache Spark

Primary LanguagePythonApache License 2.0Apache-2.0

pandas API on Apache Spark
Explore Koalas docs »

Live notebook · Issues · Mailing list
Help Thirsty Koalas Devastated by Recent Fires

The Koalas project makes data scientists more productive when interacting with big data, by implementing the pandas DataFrame API on top of Apache Spark.

pandas is the de facto standard (single-node) DataFrame implementation in Python, while Spark is the de facto standard for big data processing. With this package, you can:

  • Be immediately productive with Spark, with no learning curve, if you are already familiar with pandas.
  • Have a single codebase that works both with pandas (tests, smaller datasets) and with Spark (distributed datasets).

We would love to have you try it and give us feedback, through our mailing lists or GitHub issues.

Try the Koalas 10 minutes tutorial on a live Jupyter notebook here. The initial launch can take up to several minutes.

Github Actions codecov Documentation Status Latest Release Conda Version Binder

Getting Started

Koalas can be installed in many ways such as Conda and pip.

# Conda
conda install koalas -c conda-forge
# pip
pip install koalas

See Installation for more details.

If you are a Databricks Runtime user, you can install Koalas using the Libraries tab on the cluster UI, or using dbutils in a notebook as below for the regular Databricks Runtime,

dbutils.library.installPyPI("koalas")
dbutils.library.restartPython()

For Databricks Runtime for Machine Learning 6.0 and above, you can install it as follows.

%sh
pip install koalas

Note that Koalas requires Databricks Runtime 5.x or above. In the future, we will package Koalas out-of-the-box in both the regular Databricks Runtime and Databricks Runtime for Machine Learning.

Lastly, if your PyArrow version is 0.15+ and your PySpark version is lower than 3.0, it is best for you to set ARROW_PRE_0_15_IPC_FORMAT environment variable to 1 manually. Koalas will try its best to set it for you but it is impossible to set it if there is a Spark context already launched.

Now you can turn a pandas DataFrame into a Koalas DataFrame that is API-compliant with the former:

import databricks.koalas as ks
import pandas as pd

pdf = pd.DataFrame({'x':range(3), 'y':['a','b','b'], 'z':['a','b','b']})

# Create a Koalas DataFrame from pandas DataFrame
df = ks.from_pandas(pdf)

# Rename the columns
df.columns = ['x', 'y', 'z1']

# Do some operations in place:
df['x2'] = df.x * df.x

For more details, see Getting Started and Dependencies in the official documentation.

Contributing Guide

See Contributing Guide and Design Principles in the official documentation.

FAQ

See FAQ in the official documentation.

Best Practices

See Best Practices in the official documentation.

Koalas Talks and Blogs

See Koalas Talks and Blogs in the official documentation.