evolvr's Introduction

Evolvr

Introduction

The evolvr project goal is to create a machine learning package which utilizes

various general machine learning techniques
concurrency

Running Code

to build:
```
     mix escript.build
```
to run:
```
     ./evolvr --input=input
```
to run interactively,
```
   iex
```

Load libraries

     iex(1)> c("parallel.ex")

     iex(2)> c("maths.ex")

Now can use functions

     Evolvr.Maths.factorial(5)

     Evolvr.Maths.factorial(1000)

Running Tests

type:

       mix test

this will run both unit tests and doctests.

Code

see lib directory, eg

  lib/evolvr/maths.ex

TODO

maths library
- transpose
- matrix multiplication
- inverse
- linear solver, Ax=b
machine learning library
- linear model
- lasso
- ridge regression
- KNN
- k-means
- HMM
utilities -plotting interface eg to R
basic bioinformatic techniques
concurrency

License

see file:

license.txt

Documentation

Equations for project are located in

    doc/evolvr_equations.tex

Instructions for compling are inside tex file.

To generate project documentation with:

       mix docs

Notebook

as root:

dnf install npm

as normal user:

virtualenv -p /usr/bin/python3 notes

cd notes

source ./bin/activate

git clone https://github.com/jupyter/notebook.git

cd notebook

../bin/pip install --pre -e .

../bin/pip install jupyter-console

git clone https://github.com/pprzetacznik/IElixir.git

cd IElixir

mix deps.get

mix test

MIX_ENV=prod mix compile

mkdir -p ~/.ipython/kernels/ielixir

./install_script.sh

 jupyter notebook resources/example.ipynb

Recommend Projects