Skip to content

Latest commit

 

History

History
93 lines (58 loc) · 2.55 KB

README.rst

File metadata and controls

93 lines (58 loc) · 2.55 KB

pandas-validator

https://travis-ci.org/c-bata/pandas-validator.svg?branch=master Coveralls Status

Validates the pandas object such as DataFrame and Series. And this can define validator like django form class.

Why bugs occur in Data Wrangling with pandas

When we wrangle our data with pandas, We use DataFrame frequently. DataFrame is very powerfull and easy to handle. But DataFrame has no it's schema, so It allows irregular values without being aware of it. We are confused by these values and affect the results of data wrangling.

pandas-validator offers the functions for validating DataFrame or Series objects.

Overview

import pandas as pd
import pandas_validator as pv

class SampleDataFrameValidator(pv.DataFrameValidator):
    row_num = 5
    column_num = 2
    label1 = pv.IntegerColumnValidator('label1', min_value=0, max_value=10)
    label2 = pv.FloatColumnValidator('label2', min_value=0, max_value=10)

validator = SampleDataFrameValidator()

df = pd.DataFrame({'label1': [0, 1, 2, 3, 4], 'label2': [5.0, 6.0, 7.0, 8.0, 9.0]})
validator.is_valid(df)  # True.

df = pd.DataFrame({'label1': [11, 12, 13, 14, 15], 'label2': [5.0, 6.0, 7.0, 8.0, 9.0]})
validator.is_valid(df)  # False.

df = pd.DataFrame({'label1': [0, 1, 2], 'label2': [5.0, 6.0, 7.0]})
validator.is_valid(df)  # False

Getting Started

Requirements

  • Support python version: 2.7, 3.4, 3.5, 3.6
  • Support pandas version: 0.18, 0.19

Installation

$ pip install pandas_validator

Usage

Please see the following demo written by ipython notebook.

License

This software is licensed under the MIT License.

Resources