/apyori

A simple implementation of Apriori algorithm by Python.

Primary LanguagePythonMIT LicenseMIT

Apyori

Apyori is a simple implementation of Apriori algorithm with Python 2.7 and 3.3 - 3.5, provided as APIs and as commandline interfaces.

https://travis-ci.org/ymoch/apyori.svg?branch=master https://coveralls.io/repos/github/ymoch/apyori/badge.svg?branch=master

Module Features

  • Consisted of only one file and depends on no other libraries, which enable you to use it portably.
  • Able to used as APIs.

Application Features

  • Supports a JSON output format.
  • Supports a TSV output format for 2-items relations.

Installation

Choose one from the following.

  • Install with pip pip install apyori.
  • Put apyori.py into your project.
  • Run python setup.py install.

API Usage

Here is a basic example:

from apyori import apriori

transactions = [
    ['beer', 'nuts'],
    ['beer', 'cheese'],
]
results = list(apriori(transactions))

For more details, see apyori.apriori pydoc.

CLI Usage

First, prepare input data as tab-separated transactions.

  • Each item is separated with a tab.
  • Each transactions is separated with a line feed code.

Second, run the application. Input data is given as a standard input or file paths.

  • Run with python apyori.py command.
  • If installed, you can also run with apyori-run command.

For more details, use '-h' option.

Samples

Basic usage

apyori-run < data/integration_test_input_1.tsv

Use TSV output

apyori-run -f tsv < data/integration_test_input_1.tsv

Fields of output mean:

  • Base item.
  • Appended item.
  • Support.
  • Confidence.
  • Lift.

Specify the minimum support

apyori-run -s 0.5 < data/integration_test_input_1.tsv

Specify the minimum confidence

apyori-run -c 0.5 < data/integration_test_input_1.tsv