Skip to main content

a data typing library for machine learning

Project description

Woodwork

Tests Documentation Status PyPI Version Anaconda Version PyPI Downloads


Woodwork provides a common typing namespace for using your existing DataFrames in Featuretools, EvalML, and general ML. A Woodwork DataFrame stores the physical, logical, and semantic data types present in the data. In addition, it can store metadata about the data, allowing you to store specific information you might need for your application.

Installation

Install with pip:

python -m pip install woodwork

or from the conda-forge channel on conda:

conda install -c conda-forge woodwork

Add-ons

Update checker - Receive automatic notifications of new Woodwork releases

python -m pip install "woodwork[updater]"

Example

Below is an example of using Woodwork. In this example, a sample dataset of order items is used to create a Woodwork DataFrame, specifying the LogicalType for five of the columns.

import pandas as pd
import woodwork as ww

df = pd.read_csv("https://oss.alteryx.com/datasets/online-retail-logs-2018-08-28.csv")
df.ww.init(name='retail')
df.ww.set_types(logical_types={
    'quantity': 'Integer',
    'customer_name': 'PersonFullName',
    'country': 'Categorical',
    'order_id': 'Categorical',
    'description': 'NaturalLanguage',
})
df.ww
                   Physical Type     Logical Type Semantic Tag(s)
Column
order_id                category      Categorical    ['category']
product_id              category      Categorical    ['category']
description               string  NaturalLanguage              []
quantity                   int64          Integer     ['numeric']
order_date        datetime64[ns]         Datetime              []
unit_price               float64           Double     ['numeric']
customer_name             string   PersonFullName              []
country                 category      Categorical    ['category']
total                    float64           Double     ['numeric']
cancelled                   bool          Boolean              []

We now have initialized Woodwork on the DataFrame with the specified logical types assigned. For columns that did not have a specified logical type value, Woodwork has automatically inferred the logical type based on the underlying data. Additionally, Woodwork has automatically assigned semantic tags to some of the columns, based on the inferred or assigned logical type.

If we wanted to do further analysis on only the columns in this table that have a logical type of Boolean or a semantic tag of numeric we can simply select those columns and access a dataframe containing just those columns:

filtered_df = df.ww.select(include=['Boolean', 'numeric'])
filtered_df
    quantity  unit_price   total  cancelled
0          6      4.2075  25.245      False
1          6      5.5935  33.561      False
2          8      4.5375  36.300      False
3          6      5.5935  33.561      False
4          6      5.5935  33.561      False
..       ...         ...     ...        ...
95         6      4.2075  25.245      False
96       120      0.6930  83.160      False
97        24      0.9075  21.780      False
98        24      0.9075  21.780      False
99        24      0.9075  21.780      False

As you can see, Woodwork makes it easy to manage typing information for your data, and provides simple interfaces to access only the data you need based on the logical types or semantic tags. Please refer to the Woodwork documentation for more detail on working with a Woodwork DataFrame.

Support

The Woodwork community is happy to provide support to users of Woodwork. Project support can be found in four places depending on the type of question:

  1. For usage questions, use Stack Overflow with the woodwork tag.
  2. For bugs, issues, or feature requests start a Github issue.
  3. For discussion regarding development on the core library, use Slack.
  4. For everything else, the core developers can be reached by email at open_source_support@alteryx.com

Built at Alteryx

Woodwork is an open source project built by Alteryx. To see the other open source projects we’re working on visit Alteryx Open Source. If building impactful data science pipelines is important to you or your business, please get in touch.

Alteryx Open Source

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

woodwork-0.29.0.tar.gz (192.8 kB view hashes)

Uploaded Source

Built Distribution

woodwork-0.29.0-py3-none-any.whl (230.8 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page