SML - Standard Machine Learning Language

SML aims to proved a universal language agnostic framework to simplify the development of machine learning pipelines

Setup

Begin by cloning this repository with the following terminal command

git clone https://github.com/UI-DataScience/sml.git

Change directories into the main directory and run the following terminal command

sudo python3 setup.py develop

After running this command, the SML library will be accessible anywhere on your machine via Python.

Simple Code Example

import sml
from sml import execute

query1 = 'READ "data/auto.csv" (separator = "\s+", header = None) AND REPLACE ("?", "mode") AND \
SPLIT (train = .8, test = .2, validation = .0) AND \
REGRESS (predictors = [2,3,4,5,6,7,8], label = 1, algorithm = simple) AND \
SAVE "auto.sml"'

execute(query2, verbose=True)

With verbose = true, a “pretty” output should be printed out

Sml Summary:
=============================================
=============================================
   Dataset:        data/auto.csv
   Delimiter:      \s+
   Training Set Split:       80.00%
   Testing Set Split:        20.00%
   Predictiors:        ['2', '3', '4', '5', '6', '7', '8']
   Label:         1
   Algorithm:     simple
=============================================
=============================================

Otherwise a general output summary will be printed out

Using simple Algorithm, the dataset is from: data/auto.csv.
Currently using Predictors from column(s) ['2', '3', '4',
'5', '6', '7', '8'] and Label(s) from column(s) 1.

SML Tutorials

For detailed tutorials of SML, take a look at the SML tutorials.

Parser Documentation

For more extensive documentation on the language, take a look at the documentation for the parser

Implementation Documentation

For a more extensive documentation on the implementation details take a look at the SML documentation