Skip to main content

Library for data analysis - extracting, storing and retrieving features

Project description

mldatalib

mldatalib (Machine Learning Data Library) provides a Python library which simplifies processing and extracting features (for machine learning) from files. Stores features in a SQLite database, has label transformation options, functions which convert features to NumPy arrays, etc. Original idea and feature list by Viktor Evstratov (viktor.evst@gmail.com). Originally designed for use with the Galaxy Zoo challenge on Kaggle. Requires numpy and SQLAlchemy.

Why?

This is an attempt to minimize the amount of effort needed to extract, save and retrieve new features and allow a user to spend more time on more ‘scientific’ work. If several users are working on the same project, they can each extract independent sets of features and then share the database files and copy the features they are missing.

Functionality

Basic functionality includes: Extracting features and storing them in a database (the user provides the extractor function), retrieving features by name, extracting and transforming labels from a file and storing them in a database, copying features from one database to another, returning features as a numpy array.

Roadmap

Add a pure SQL way of storing and retrieving features (by converting them to JSON format and creating columns via ALTER TABLE statements), thus allowing easy use from other languages. At least add this functionality for numpy arrays and lists.

Add a dataset class for CSV files (where all data is stored in a single CSV file).

Project details


Release history Release notifications

History Node

0.2.1

History Node

0.2

History Node

0.1.4

History Node

0.1.3

This version
History Node

0.1.2

History Node

0.1.1

History Node

0.1

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
mldatalib-0.1.2.tar.gz (11.6 kB) Copy SHA256 hash SHA256 Source None Mar 19, 2014

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging CloudAMQP CloudAMQP RabbitMQ AWS AWS Cloud computing Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page