https://github.com/dusenberrymw/pine

A simple neural net library written entirely in vanilla Python 3.
https://github.com/dusenberrymw/pine

Last synced: 2 months ago
JSON representation

A simple neural net library written entirely in vanilla Python 3.

• Host: GitHub
• URL: https://github.com/dusenberrymw/pine
• Owner: dusenberrymw
• License: mit
• Created: 2013-09-10T23:47:13.000Z (over 11 years ago)
• Default Branch: master
• Last Pushed: 2015-02-03T20:57:29.000Z (almost 10 years ago)
• Last Synced: 2023-08-03T19:58:05.773Z (over 1 year ago)
• Language: Python
• Homepage:
• Size: 470 KB
• Stars: 5
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Pine: Python Neural Networks

---

#### A simple ANN library written entirely in vanilla Python 3

This has been a fun project that was started in order to further my understanding of neural nets and machine learning in general.  It's also being used for research in Emergency Medicine.

## Requirements

Python3 installed (and accessible at `/usr/bin/env python3`)

 

## Library

View the `pine/` directory -> everything should be pretty well documented

 

Use `python3 setup.py install` to install globally

 

## Command line script

Use `./bin/pineCLI.py` (add `-h` for help)

 

##### Install Shell Command:

To access pineCLI.py globally, run `./installShellCommand.py`.  This will install `pine` to /usr/local/bin/ as a soft link, allowing `bin/pineCLI.py` to be run as `pine` from any directory, assuming /usr/local/bin/ is on your path.

  

## Demo

Use `./demo/demo.py` for a quick, old demo.  Also, the bottom of demo.py can be edited to run other sample demo projects.  

## Workflow

Checkout the [Pine Data Tools](https://github.com/dusenberrymw/Pine-Data-Tools) repo for a sample workflow for running real projects.  Be sure to install the shell command first, as seen above.