Machine Learning for Go
Clone or download
Pull request Compare This branch is 83 commits behind sjwhitworth:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
base
clustering
doc
ensemble
evaluation
examples
filters
kdtree
knn
linear_models
meta
metrics/pairwise
naive
neural
optimisation
pca
perceptron
trees
utilities
.gitignore
.travis.yml
Dockerfile
LICENSE.md
README.md
golearn.go

README.md

GoLearn


GoDoc Build Status

Support via Gittip

GoLearn is a 'batteries included' machine learning library for Go. Simplicity, paired with customisability, is the goal. We are in active development, and would love comments from users out in the wild. Drop us a line on Twitter.

twitter: @golearn_ml

Install

See here for installation instructions.

Getting Started

Data are loaded in as Instances. You can then perform matrix like operations on them, and pass them to estimators. GoLearn implements the scikit-learn interface of Fit/Predict, so you can easily swap out estimators for trial and error. GoLearn also includes helper functions for data, like cross validation, and train and test splitting.

package main

import (
	"fmt"

	"github.com/sjwhitworth/golearn/base"
	"github.com/sjwhitworth/golearn/evaluation"
	"github.com/sjwhitworth/golearn/knn"
)

func main() {
	// Load in a dataset, with headers. Header attributes will be stored.
	// Think of instances as a Data Frame structure in R or Pandas.
	// You can also create instances from scratch.
	rawData, err := base.ParseCSVToInstances("datasets/iris.csv", false)
	if err != nil {
		panic(err)
	}

	// Print a pleasant summary of your data.
	fmt.Println(rawData)

	//Initialises a new KNN classifier
	cls := knn.NewKnnClassifier("euclidean", "linear", 2)

	//Do a training-test split
	trainData, testData := base.InstancesTrainTestSplit(rawData, 0.50)
	cls.Fit(trainData)

	//Calculates the Euclidean distance and returns the most popular label
	predictions, err := cls.Predict(testData)
	if err != nil {
		panic(err)
	}

	// Prints precision/recall metrics
	confusionMat, err := evaluation.GetConfusionMatrix(testData, predictions)
	if err != nil {
		panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
	}
	fmt.Println(evaluation.GetSummary(confusionMat))
}
Iris-virginica	28	2	  56	0.9333	0.9333  0.9333
Iris-setosa	    29	0	  59	1.0000  1.0000	1.0000
Iris-versicolor	27	2	  57	0.9310	0.9310  0.9310
Overall accuracy: 0.9545

Examples

GoLearn comes with practical examples. Dive in and see what is going on.

cd $GOPATH/src/github.com/sjwhitworth/golearn/examples/knnclassifier
go run knnclassifier_iris.go
cd $GOPATH/src/github.com/sjwhitworth/golearn/examples/instances
go run instances.go
cd $GOPATH/src/github.com/sjwhitworth/golearn/examples/trees
go run trees.go

Docs

Join the team

Please send me a mail at stephenjameswhitworth@gmail.com