User Guide

Diogenes is a set of tools designed to make common machine learning tasks easier. Diogenes is divided into several parts:

• The Read Module provides tools for reading data from external sources into Diogenes’ preferred Numpy structured array format.
• The Display Module provides tools for summarizing data and the performance of trained classifiers.
• The Modify Module provides tools for manipulating arrays and generating features.
• The Grid Search Module provides tools for finding the best classifier, testing classifier sensitivity to data sets, and cross-validating classifier performance.
• The Array Emitter Module provides tools for processing and iterating over “RG-formatted” (transposed) arrays.
• The Utils Module provides miscilaneous utilities–mostly for processing structured arrays.

Contents:

• The Read Module
  • Local CSV Files
  • Remote CSV files
  • SQL databases
• The Display Module
  • Exploring data
  • Examining classifier performance.
  • Making PDF Reports
• The Modify Module
  • In-place Cleaning
  • Selection
  • Feature Generation
  • Last steps
• The Grid Search Module
  • Experiments, Classifiers and Hyperparameters
  • Subset Iterators and Partition Iterators
  • Trials and Runs
  • Reports
• The Array Emitter Module
• The Utils Module