SPMF

Python Wrapper for SPMF Java library.

Information

This module contains python wrappers for pattern mining algorithms implemented in SPMF Java library. Each algorithm is implemented as a standalone Python class with fully descriptive and tested APIs. It also provides native support for Pandas dataframes.

Why? If you're in a Python pipeline, it might be cumbersome to use Java as an intermediate step. Using spmf-wrapper you can stay in your pipeline as though Java is never used at all.

Installation

pip install spmf-wrapper

A Java Runtime Environment is required to run this wrapper. If an existing installation is not detected, JRE v21 is automatically installed using install-jdk python module at $HOME/.jre/jdk-21.0.2+13-jre. If you prefer to install Java Runtime manually, follow instructions here. Test installation by running the following command on the terminal:

> java -version
java version "1.8.0_391"
Java(TM) SE Runtime Environment (build 1.8.0_391-b13)
Java HotSpot(TM) 64-Bit Server VM (build 25.391-b13, mixed mode)

Usage

Example:

from spmf import EMMA

emma = EMMA(min_support=2, max_window=2, timestamp_present=True, transform=True)
output = emma.run_pandas(input_df)

Input:

Output:

See examples for more details.

For a detailed explanation of the algorithm and parameters, refer to the corresponding webpage in the SPMF documentation.

Implementation Checklist

Sequential Pattern Mining

Episode Mining

Bibliography

Fournier-Viger, P., Lin, C.W., Gomariz, A., Gueniche, T., Soltani, A., Deng, Z., Lam, H. T. (2016).
The SPMF Open-Source Data Mining Library Version 2.
Proc. 19th European Conference on Principles of Data Mining and Knowledge Discovery (PKDD 2016) Part III, Springer LNCS 9853,  pp. 36-40.