---

Documentation | Forum | Mail List

Dataprep lets you prepare your data using a single library with a few lines of code.

Currently, you can use dataprep to:

• Collect data from common data sources (through dataprep.connector)
• Do your exploratory data analysis (through dataprep.eda)
• ...more modules are coming

Releases

Repo	Version	Downloads
PyPI
conda-forge

Installation

pip install -U dataprep

Examples & Usages

The following examples can give you an impression of what dataprep can do:

• Documentation: Connector
• Documentation: EDA
• EDA Case Study: Titanic
• EDA Case Study: House Price

EDA

There are common tasks during the exploratory data analysis stage, like a quick look at the columnar distribution, or understanding the correlations between columns.

The EDA module categorizes these EDA tasks into functions helping you finish EDA tasks with a single function call.

• Want to understand the distributions for each DataFrame column? Use plot.

• Want to understand the correlation between columns? Use plot_correlation.

• Or, if you want to understand the impact of the missing values for each column, use plot_missing.

You can drill down to get more information by given plot, plot_correlation and plot_missing a column name.: E.g. for plot_missing

    for numerical column usingplot:

    for categorical column usingplot:

Don't forget to checkout the examples folder for detailed demonstration!

Connector

Connector provides a simple way to collect data from different websites, offering several benefits:

• A unified API: you can fetch data using one or two lines of code to get data from many websites.
• Auto Pagination: it automatically does the pagination for you so that you can specify the desired count of the returned results without even considering the count-per-request restriction from the API.
• Smart API request strategy: it can issue API requests in parallel while respecting the rate limit policy.

In the following examples, you can download the Yelp business search result into a pandas DataFrame, using only two lines of code, without taking deep looking into the Yelp documentation! More examples can be found here: Examples

Clean

DataPrep.Clean contains simple functions designed for cleaning and standardizing a column in a DataFrame. It provides

• A unified API: each function follows the syntax clean_{type}(df, "column name") (see an example below)
• Python Data Science Support: its design for cleaning pandas and Dask DataFrames enables seamless integration into the Python data science workflow
• Transparency: a report is generated that summarizes the alterations to the data that occured during cleaning

The following example shows how to clean a column containing messy emails:

Type validation is also supported:

Below are the supported semantic types (more are currently being developed).

Semantic Types
longitude/latitude
country
email
url
phone

For more information, refer to the User Guide.

Contribute

There are many ways to contribute to Dataprep.

• Submit bugs and help us verify fixes as they are checked in.
• Review the source code changes.
• Engage with other Dataprep users and developers on StackOverflow.
• Help each other in the Dataprep Community Discord and Mail list & Forum.
• 
• Contribute bug fixes.
• Providing use cases and writing down your user experience.

Please take a look at our wiki for development documentations!

Acknowledgement

Some functionalities of DataPrep are inspired by the following packages.

• Pandas Profiling

  Inspired the report functionality and insights provided in DataPrep.eda.

• missingno

  Inspired the missing value analysis in DataPrep.eda.