canomaly 0.0.5

Detecting cumulative changes in data.

canomaly

Project Description

This package detects specific types of anomalies with an emphasis in looking for cumulative changes.

Installation

This package can be installed through PyPi using

pip install canomaly

or

pip3 install canomaly

Example Usage

>>> import pandas as pd
>>> from canomaly.searchtools import cumrexpy
>>> # Get some data
>>> data = {
            'date': [
                '2018-11-20',
                '2018-11-21',
                '2018-11-22',
                '2018-11-22',
                '2018-11-23',
                '2018-11-24'],
            'email': [
                '[e-mail]',
                '[e-mail]',
                '[e-mail]',
                '[e-mail]',
                '[e-mail]',
                '[e-mail]'
                    ]
            }
>>> df = pd.DataFrame(data)
>>> df['date'] = pd.to_datetime(df['date'])
>>> # Extract regular expressions
>>> cumrexpy(df, 'email', 'date')
date
2018-11-20                           [^john\.doe@example\.com$]
2018-11-21                [^[a-z]{4}\.[a-z]{3,5}@example\.com$]
2018-11-22    [^[a-z]{4,5}[.@][a-z]+[.@][a-z]+\.[a-z]{2,3}$,...
2018-11-23    [^frank@mydomain\.com$, ^[a-z]{4,5}[.@][a-z]+[...
2018-11-24    [^frank@mydomain\.com$, ^[a-z]+[.@_][a-z]+[.@]...
Name: email_grouped, dtype: object

We can look at the results in markdown for clarity.

date	email_grouped
2018-11-20 00:00:00	['^john\.doe@example\.com$']
2018-11-21 00:00:00	['^[a-z]{4}\.[a-z]{3,5}@example\.com$']
2018-11-22 00:00:00	['^[a-z]{4,5}[.@][a-z]+[.@][a-z]+\.[a-z]{2,3}$', '^bob\-johnson_123@example\.com$']
2018-11-23 00:00:00	['^frank@mydomain\.com$', '^[a-z]{4,5}[.@][a-z]+[.@][a-z]+\.[a-z]{2,3}$', '^bob\-johnson_123@example\.com$']
2018-11-24 00:00:00	['^frank@mydomain\.com$', '^[a-z]+[.@_][a-z]+[.@][a-z]+\.[a-z]{2,3}$', '^bob\-johnson_123@example\.com$']

Build Documentation Locally

cd /path/to/canomaly/docs
make html