data-quality-tests 0.1.5.3

Data Quality Check Library

DATA QUALITY

A library which acts as a test cases for dataframes. Simply pass in your dataframe after initial import, or at each stage of your EDA to check for data quality with one line of code.

The test cases include(as of now)

1. check for null values
2. check for duplicates
3. check for dtype matching
4. check for outliers
5. check for whitespaces in column headers

The test cases work as a Pass/Fail type, where Passed indicates, good data quality and Failed indicates bad data quality

Example:

TEST CASE FOR NULL VALUES: Passed means that the dataframe has no null values. Failed indicates otherwise.

Requirements

• Python 3+
• Pandas
• Numpy

Installation

pip install data-quality-tests

Updates & Changes

1. the import function changed from:

from data_quality import DataQuality

to the following:

from data_quality_tests import DataQuality

2. new function outlier_columns has been added in this update, which displays all the columns that have outliers.
   For use case, refer to the get started section

3. data_quality_check now checks for column header whitespaces for leading and trailing.

Get Started

How to use this library:

Data quality check

The most basic usage of this library, here for simplifiction,
let's just se the iris dataset from seaborn library.

You can use any dataset.

from data_quality_tests import DataQuality as dq
import seaborn as sns

#declare any dataframe

df = sns.load_dataset("iris")

#pass the dataframe as below  

dq.data_quality_check(df)

Outlier columns

Sometimes, the test case for outliers fails, this is because the dataset containes outliers.

use outier_columns(df) function to display all the columns that have outliers.

NOTE If the dataset does not have outlier columns, the output is an empty list.

# display columns that have outliers
dq.outlier_columns(df)