func main() {
var tree base.Classifier
rand.Seed(time.Now().UTC().UnixNano())
// Load in the iris dataset
iris, err := base.ParseCSVToInstances("../datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
// Discretise the iris dataset with Chi-Merge
filt := filters.NewChiMergeFilter(iris, 0.99)
filt.AddAllNumericAttributes()
filt.Build()
filt.Run(iris)
// Create a 60-40 training-test split
insts := base.InstancesTrainTestSplit(iris, 0.60)
//
// First up, use ID3
//
tree = trees.NewID3DecisionTree(0.6)
// (Parameter controls train-prune split.)
// Train the ID3 tree
tree.Fit(insts[0])
// Generate predictions
predictions := tree.Predict(insts[1])
// Evaluate
fmt.Println("ID3 Performance")
cf := eval.GetConfusionMatrix(insts[1], predictions)
fmt.Println(eval.GetSummary(cf))
//
// Next up, Random Trees
//
// Consider two randomly-chosen attributes
tree = trees.NewRandomTree(2)
tree.Fit(insts[0])
predictions = tree.Predict(insts[1])
fmt.Println("RandomTree Performance")
cf = eval.GetConfusionMatrix(insts[1], predictions)
fmt.Println(eval.GetSummary(cf))
//
// Finally, Random Forests
//
tree = ensemble.NewRandomForest(100, 3)
tree.Fit(insts[0])
predictions = tree.Predict(insts[1])
fmt.Println("RandomForest Performance")
cf = eval.GetConfusionMatrix(insts[1], predictions)
fmt.Println(eval.GetSummary(cf))
}
func TestRandomForest1(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
rand.Seed(time.Now().UnixNano())
trainData, testData := base.InstancesTrainTestSplit(inst, 0.6)
filt := filters.NewChiMergeFilter(inst, 0.90)
for _, a := range base.NonClassFloatAttributes(inst) {
filt.AddAttribute(a)
}
filt.Train()
trainDataf := base.NewLazilyFilteredInstances(trainData, filt)
testDataf := base.NewLazilyFilteredInstances(testData, filt)
rf := new(BaggedModel)
for i := 0; i < 10; i++ {
rf.AddModel(trees.NewRandomTree(2))
}
rf.Fit(trainDataf)
fmt.Println(rf)
predictions := rf.Predict(testDataf)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testDataf, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
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", 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 := cls.Predict(testData)
fmt.Println(predictions)
// 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))
}
func TestPruning(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
trainData, testData := base.InstancesTrainTestSplit(inst, 0.6)
filt := filters.NewChiMergeFilter(inst, 0.90)
filt.AddAllNumericAttributes()
filt.Build()
fmt.Println(testData)
filt.Run(testData)
filt.Run(trainData)
root := NewRandomTree(2)
fittrainData, fittestData := base.InstancesTrainTestSplit(trainData, 0.6)
root.Fit(fittrainData)
root.Prune(fittestData)
fmt.Println(root)
predictions := root.Predict(testData)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testData, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
func TestRandomForest1(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
rand.Seed(time.Now().UnixNano())
insts := base.InstancesTrainTestSplit(inst, 0.6)
filt := filters.NewChiMergeFilter(inst, 0.90)
filt.AddAllNumericAttributes()
filt.Build()
filt.Run(insts[1])
filt.Run(insts[0])
rf := new(BaggedModel)
for i := 0; i < 10; i++ {
rf.AddModel(trees.NewRandomTree(2))
}
rf.Fit(insts[0])
fmt.Println(rf)
predictions := rf.Predict(insts[1])
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(insts[1], predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
func TestPredict(t *testing.T) {
a := NewAveragePerceptron(10, 1.2, 0.5, 0.3)
if a == nil {
t.Errorf("Unable to create average perceptron")
}
absPath, _ := filepath.Abs("../examples/datasets/house-votes-84.csv")
rawData, err := base.ParseCSVToInstances(absPath, true)
if err != nil {
t.Fail()
}
trainData, testData := base.InstancesTrainTestSplit(rawData, 0.5)
a.Fit(trainData)
if a.trained == false {
t.Errorf("Perceptron was not trained")
}
predictions := a.Predict(testData)
cf, err := evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
t.Errorf("Couldn't get confusion matrix: %s", err)
t.Fail()
}
fmt.Println(evaluation.GetSummary(cf))
fmt.Println(trainData)
fmt.Println(testData)
if evaluation.GetAccuracy(cf) < 0.65 {
t.Errorf("Perceptron not trained correctly")
}
}
func BenchmarkKNNWithNoOpts(b *testing.B) {
// Load
train, test := readMnist()
cls := NewKnnClassifier("euclidean", 1)
cls.AllowOptimisations = false
cls.Fit(train)
predictions := cls.Predict(test)
c, err := evaluation.GetConfusionMatrix(test, predictions)
if err != nil {
panic(err)
}
fmt.Println(evaluation.GetSummary(c))
fmt.Println(evaluation.GetAccuracy(c))
}
func TestMultiSVMUnweighted(t *testing.T) {
Convey("Loading data...", t, func() {
inst, err := base.ParseCSVToInstances("../examples/datasets/articles.csv", false)
So(err, ShouldBeNil)
X, Y := base.InstancesTrainTestSplit(inst, 0.4)
m := NewMultiLinearSVC("l1", "l2", true, 1.0, 1e-4, nil)
m.Fit(X)
Convey("Predictions should work...", func() {
predictions, err := m.Predict(Y)
cf, err := evaluation.GetConfusionMatrix(Y, predictions)
So(err, ShouldEqual, nil)
So(evaluation.GetAccuracy(cf), ShouldBeGreaterThan, 0.70)
})
})
}
func main() {
data, err := base.ParseCSVToInstances("iris_headers.csv", true)
if err != nil {
panic(err)
}
cls := knn.NewKnnClassifier("euclidean", 2)
trainData, testData := base.InstancesTrainTestSplit(data, 0.8)
cls.Fit(trainData)
predictions := cls.Predict(testData)
fmt.Println(predictions)
confusionMat := evaluation.GetConfusionMatrix(testData, predictions)
fmt.Println(evaluation.GetSummary(confusionMat))
}
func NewTestTrial(filename string, split float64) bool {
cls := knn.NewKnnClassifier("euclidean", 2)
data := CSVtoKNNData(filename)
train, test := base.InstancesTrainTestSplit(data, split)
cls.Fit(train)
//Calculates the Euclidean distance and returns the most popular label
predictions := cls.Predict(test)
fmt.Println(predictions)
confusionMat, err := evaluation.GetConfusionMatrix(test, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(confusionMat))
return true
}
func TestRandomForest(t *testing.T) {
Convey("Given a valid CSV file", t, func() {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
So(err, ShouldBeNil)
Convey("When Chi-Merge filtering the data", func() {
filt := filters.NewChiMergeFilter(inst, 0.90)
for _, a := range base.NonClassFloatAttributes(inst) {
filt.AddAttribute(a)
}
filt.Train()
instf := base.NewLazilyFilteredInstances(inst, filt)
Convey("Splitting the data into test and training sets", func() {
trainData, testData := base.InstancesTrainTestSplit(instf, 0.60)
Convey("Fitting and predicting with a Random Forest", func() {
rf := NewRandomForest(10, 3)
err = rf.Fit(trainData)
So(err, ShouldBeNil)
predictions, err := rf.Predict(testData)
So(err, ShouldBeNil)
confusionMat, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMat), ShouldBeGreaterThan, 0.35)
})
})
})
})
Convey("Fitting with a Random Forest with too many features compared to the data", func() {
rf := NewRandomForest(10, len(base.NonClassAttributes(inst))+1)
err = rf.Fit(inst)
Convey("Should return an error", func() {
So(err, ShouldNotBeNil)
})
})
})
}
func TestRandomForest1(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
trainData, testData := base.InstancesTrainTestSplit(inst, 0.60)
filt := filters.NewChiMergeFilter(trainData, 0.90)
filt.AddAllNumericAttributes()
filt.Build()
filt.Run(testData)
filt.Run(trainData)
rf := NewRandomForest(10, 3)
rf.Fit(trainData)
predictions := rf.Predict(testData)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testData, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetSummary(confusionMat))
}
func main() {
rawData, err := base.ParseCSVToInstances("../datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
//Initialises a new KNN classifier
cls := knn.NewKnnClassifier("euclidean", 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 := cls.Predict(testData)
fmt.Println(predictions)
// Prints precision/recall metrics
confusionMat := evaluation.GetConfusionMatrix(testData, predictions)
fmt.Println(evaluation.GetSummary(confusionMat))
}
func TestMultiSVMWeighted(t *testing.T) {
Convey("Loading data...", t, func() {
weights := make(map[string]float64)
weights["Finance"] = 0.1739
weights["Tech"] = 0.0750
weights["Politics"] = 0.4928
inst, err := base.ParseCSVToInstances("../examples/datasets/articles.csv", false)
So(err, ShouldBeNil)
X, Y := base.InstancesTrainTestSplit(inst, 0.4)
m := NewMultiLinearSVC("l1", "l2", true, 0.62, 1e-4, weights)
m.Fit(X)
Convey("Predictions should work...", func() {
predictions, err := m.Predict(Y)
cf, err := evaluation.GetConfusionMatrix(Y, predictions)
So(err, ShouldEqual, nil)
So(evaluation.GetAccuracy(cf), ShouldBeGreaterThan, 0.70)
})
})
}
func main() {
// Load and parse the data from csv files
fmt.Println("Loading data...")
trainData, err := base.ParseCSVToInstances("data/mnist_train.csv", true)
if err != nil {
panic(err)
}
testData, err := base.ParseCSVToInstances("data/mnist_test.csv", true)
if err != nil {
panic(err)
}
// Create a new linear SVC with some good default values
classifier, err := linear_models.NewLinearSVC("l1", "l2", true, 1.0, 1e-4)
if err != nil {
panic(err)
}
// Don't output information on each iteration
base.Silent()
// Train the linear SVC
fmt.Println("Training...")
classifier.Fit(trainData)
// Make predictions for the test data
fmt.Println("Predicting...")
predictions, err := classifier.Predict(testData)
if err != nil {
panic(err)
}
// Get a confusion matrix and print out some accuracy stats for our predictions
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))
}
func TestOneVsAllModel(t *testing.T) {
classifierFunc := func(c string) base.Classifier {
m, err := linear_models.NewLinearSVC("l1", "l2", true, 1.0, 1e-4)
if err != nil {
panic(err)
}
return m
}
Convey("Given data", t, func() {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
So(err, ShouldBeNil)
X, Y := base.InstancesTrainTestSplit(inst, 0.4)
m := NewOneVsAllModel(classifierFunc)
m.Fit(X)
Convey("The maximum class index should be 2", func() {
So(m.maxClassVal, ShouldEqual, 2)
})
Convey("There should be three of everything...", func() {
So(len(m.filters), ShouldEqual, 3)
So(len(m.classifiers), ShouldEqual, 3)
})
Convey("Predictions should work...", func() {
predictions, err := m.Predict(Y)
So(err, ShouldEqual, nil)
cf, err := evaluation.GetConfusionMatrix(Y, predictions)
So(err, ShouldEqual, nil)
fmt.Println(evaluation.GetAccuracy(cf))
fmt.Println(evaluation.GetSummary(cf))
})
})
}
func TestBaggedModelRandomForest(t *testing.T) {
Convey("Given data", t, func() {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
So(err, ShouldBeNil)
Convey("Splitting the data into training and test data", func() {
trainData, testData := base.InstancesTrainTestSplit(inst, 0.6)
Convey("Filtering the split datasets", func() {
rand.Seed(time.Now().UnixNano())
filt := filters.NewChiMergeFilter(inst, 0.90)
for _, a := range base.NonClassFloatAttributes(inst) {
filt.AddAttribute(a)
}
filt.Train()
trainDataf := base.NewLazilyFilteredInstances(trainData, filt)
testDataf := base.NewLazilyFilteredInstances(testData, filt)
Convey("Fitting and Predicting with a Bagged Model of 10 Random Trees", func() {
rf := new(BaggedModel)
for i := 0; i < 10; i++ {
rf.AddModel(trees.NewRandomTree(2))
}
rf.Fit(trainDataf)
predictions := rf.Predict(testDataf)
confusionMat, err := evaluation.GetConfusionMatrix(testDataf, predictions)
So(err, ShouldBeNil)
Convey("Predictions are somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMat), ShouldBeGreaterThan, 0.5)
})
})
})
})
})
}
func TestRandomForest1(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
filt := filters.NewChiMergeFilter(inst, 0.90)
for _, a := range base.NonClassFloatAttributes(inst) {
filt.AddAttribute(a)
}
filt.Train()
instf := base.NewLazilyFilteredInstances(inst, filt)
trainData, testData := base.InstancesTrainTestSplit(instf, 0.60)
rf := NewRandomForest(10, 3)
rf.Fit(trainData)
predictions := rf.Predict(testData)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testData, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetSummary(confusionMat))
}
func TestID3Classification(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
filt := filters.NewBinningFilter(inst, 10)
filt.AddAllNumericAttributes()
filt.Build()
filt.Run(inst)
fmt.Println(inst)
trainData, testData := base.InstancesTrainTestSplit(inst, 0.70)
// Build the decision tree
rule := new(InformationGainRuleGenerator)
root := InferID3Tree(trainData, rule)
fmt.Println(root)
predictions := root.Predict(testData)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testData, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
func TestRandomTreeClassification2(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
insts := base.InstancesTrainTestSplit(inst, 0.4)
filt := filters.NewChiMergeFilter(inst, 0.90)
filt.AddAllNumericAttributes()
filt.Build()
fmt.Println(insts[1])
filt.Run(insts[1])
filt.Run(insts[0])
root := NewRandomTree(2)
root.Fit(insts[0])
fmt.Println(root)
predictions := root.Predict(insts[1])
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(insts[1], predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
func main() {
rand.Seed(4402201)
rawData, err := base.ParseCSVToInstances("../datasets/house-votes-84.csv", true)
if err != nil {
panic(err)
}
//Initialises a new AveragePerceptron classifier
cls := perceptron.NewAveragePerceptron(10, 1.2, 0.5, 0.3)
//Do a training-test split
trainData, testData := base.InstancesTrainTestSplit(rawData, 0.50)
fmt.Println(trainData)
fmt.Println(testData)
cls.Fit(trainData)
predictions := cls.Predict(testData)
// Prints precision/recall metrics
confusionMat, _ := evaluation.GetConfusionMatrix(testData, predictions)
fmt.Println(evaluation.GetSummary(confusionMat))
}
func TestRandomTreeClassification(testEnv *testing.T) {
inst, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
trainData, testData := base.InstancesTrainTestSplit(inst, 0.6)
filt := filters.NewChiMergeFilter(inst, 0.90)
filt.AddAllNumericAttributes()
filt.Build()
filt.Run(trainData)
filt.Run(testData)
fmt.Println(inst)
r := new(RandomTreeRuleGenerator)
r.Attributes = 2
root := InferID3Tree(trainData, r)
fmt.Println(root)
predictions := root.Predict(testData)
fmt.Println(predictions)
confusionMat := eval.GetConfusionMatrix(testData, predictions)
fmt.Println(confusionMat)
fmt.Println(eval.GetMacroPrecision(confusionMat))
fmt.Println(eval.GetMacroRecall(confusionMat))
fmt.Println(eval.GetSummary(confusionMat))
}
func TestRandomTreeClassification(t *testing.T) {
Convey("Predictions on filtered data with a Random Tree", t, func() {
instances, err := base.ParseCSVToInstances("../examples/datasets/iris_headers.csv", true)
So(err, ShouldBeNil)
trainData, testData := base.InstancesTrainTestSplit(instances, 0.6)
filter := filters.NewChiMergeFilter(instances, 0.9)
for _, a := range base.NonClassFloatAttributes(instances) {
filter.AddAttribute(a)
}
filter.Train()
filteredTrainData := base.NewLazilyFilteredInstances(trainData, filter)
filteredTestData := base.NewLazilyFilteredInstances(testData, filter)
Convey("Using InferID3Tree to create the tree and do the fitting", func() {
Convey("Using a RandomTreeRule", func() {
randomTreeRuleGenerator := new(RandomTreeRuleGenerator)
randomTreeRuleGenerator.Attributes = 2
root := InferID3Tree(filteredTrainData, randomTreeRuleGenerator)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(filteredTestData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(filteredTestData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
Convey("Using a InformationGainRule", func() {
informationGainRuleGenerator := new(InformationGainRuleGenerator)
root := InferID3Tree(filteredTrainData, informationGainRuleGenerator)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(filteredTestData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(filteredTestData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
})
Convey("Using NewRandomTree to create the tree", func() {
root := NewRandomTree(2)
Convey("Fitting with the tree", func() {
err = root.Fit(filteredTrainData)
So(err, ShouldBeNil)
Convey("Predicting with the tree, *without* pruning first", func() {
predictions, err := root.Predict(filteredTestData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(filteredTestData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
Convey("Predicting with the tree, pruning first", func() {
root.Prune(filteredTestData)
predictions, err := root.Predict(filteredTestData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(filteredTestData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.4)
})
})
})
})
})
}
func verifyTreeClassification(trainData, testData base.FixedDataGrid) {
rand.Seed(44414515)
Convey("Using InferID3Tree to create the tree and do the fitting", func() {
Convey("Using a RandomTreeRule", func() {
randomTreeRuleGenerator := new(RandomTreeRuleGenerator)
randomTreeRuleGenerator.Attributes = 2
root := InferID3Tree(trainData, randomTreeRuleGenerator)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
Convey("Using a InformationGainRule", func() {
informationGainRuleGenerator := new(InformationGainRuleGenerator)
root := InferID3Tree(trainData, informationGainRuleGenerator)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
Convey("Using a GiniCoefficientRuleGenerator", func() {
gRuleGen := new(GiniCoefficientRuleGenerator)
root := InferID3Tree(trainData, gRuleGen)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
Convey("Using a InformationGainRatioRuleGenerator", func() {
gRuleGen := new(InformationGainRatioRuleGenerator)
root := InferID3Tree(trainData, gRuleGen)
Convey("Predicting with the tree", func() {
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
})
})
Convey("Using NewRandomTree to create the tree", func() {
root := NewRandomTree(2)
Convey("Fitting with the tree", func() {
err := root.Fit(trainData)
So(err, ShouldBeNil)
Convey("Predicting with the tree, *without* pruning first", func() {
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.5)
})
})
Convey("Predicting with the tree, pruning first", func() {
root.Prune(testData)
predictions, err := root.Predict(testData)
So(err, ShouldBeNil)
confusionMatrix, err := evaluation.GetConfusionMatrix(testData, predictions)
So(err, ShouldBeNil)
Convey("Predictions should be somewhat accurate", func() {
So(evaluation.GetAccuracy(confusionMatrix), ShouldBeGreaterThan, 0.4)
})
})
})
})
}
// computeAccuracy is a helper method for Prune()
func computeAccuracy(predictions base.FixedDataGrid, from base.FixedDataGrid) float64 {
cf, _ := evaluation.GetConfusionMatrix(from, predictions)
return evaluation.GetAccuracy(cf)
}
func main() {
var tree base.Classifier
rand.Seed(44111342)
// Load in the iris dataset
iris, err := base.ParseCSVToInstances("../datasets/iris_headers.csv", true)
if err != nil {
panic(err)
}
// Discretise the iris dataset with Chi-Merge
filt := filters.NewChiMergeFilter(iris, 0.999)
for _, a := range base.NonClassFloatAttributes(iris) {
filt.AddAttribute(a)
}
filt.Train()
irisf := base.NewLazilyFilteredInstances(iris, filt)
// Create a 60-40 training-test split
trainData, testData := base.InstancesTrainTestSplit(irisf, 0.60)
//
// First up, use ID3
//
tree = trees.NewID3DecisionTree(0.6)
// (Parameter controls train-prune split.)
// Train the ID3 tree
err = tree.Fit(trainData)
if err != nil {
panic(err)
}
// Generate predictions
predictions, err := tree.Predict(testData)
if err != nil {
panic(err)
}
// Evaluate
fmt.Println("ID3 Performance (information gain)")
cf, err := evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(cf))
tree = trees.NewID3DecisionTreeFromRule(0.6, new(trees.InformationGainRatioRuleGenerator))
// (Parameter controls train-prune split.)
// Train the ID3 tree
err = tree.Fit(trainData)
if err != nil {
panic(err)
}
// Generate predictions
predictions, err = tree.Predict(testData)
if err != nil {
panic(err)
}
// Evaluate
fmt.Println("ID3 Performance (information gain ratio)")
cf, err = evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(cf))
tree = trees.NewID3DecisionTreeFromRule(0.6, new(trees.GiniCoefficientRuleGenerator))
// (Parameter controls train-prune split.)
// Train the ID3 tree
err = tree.Fit(trainData)
if err != nil {
panic(err)
}
// Generate predictions
predictions, err = tree.Predict(testData)
if err != nil {
panic(err)
}
// Evaluate
fmt.Println("ID3 Performance (gini index generator)")
cf, err = evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(cf))
//
// Next up, Random Trees
//
// Consider two randomly-chosen attributes
tree = trees.NewRandomTree(2)
err = tree.Fit(testData)
if err != nil {
panic(err)
}
predictions, err = tree.Predict(testData)
if err != nil {
panic(err)
}
fmt.Println("RandomTree Performance")
cf, err = evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(cf))
//
// Finally, Random Forests
//
tree = ensemble.NewRandomForest(70, 3)
err = tree.Fit(trainData)
if err != nil {
panic(err)
}
predictions, err = tree.Predict(testData)
if err != nil {
panic(err)
}
fmt.Println("RandomForest Performance")
cf, err = evaluation.GetConfusionMatrix(testData, predictions)
if err != nil {
panic(fmt.Sprintf("Unable to get confusion matrix: %s", err.Error()))
}
fmt.Println(evaluation.GetSummary(cf))
}
// computeAccuracy is a helper method for Prune()
func computeAccuracy(predictions *base.Instances, from *base.Instances) float64 {
cf := eval.GetConfusionMatrix(from, predictions)
return eval.GetAccuracy(cf)
}