Documentation
¶
Overview ¶
Package utils provides utility functions for outlier detection.
Package utils provides utility functions for outlier detection.
Package utils provides utility functions for outlier detection.
Index ¶
- Variables
- func Accuracy(yTrue, yPred []int) float64
- func ApplyStandardize(X [][]float64, means, stds []float64) [][]float64
- func ArgMaxN(values []float64, n int) []int
- func ArgMinN(values []float64, n int) []int
- func BalancedAccuracy(yTrue, yPred []int) float64
- func CheckParameter(param, low, high float64, includeLow, includeHigh bool) error
- func ConfusionMatrix(yTrue, yPred []int) [4]int
- func F1Score(yTrue, yPred []int) float64
- func GenerateData(opts *GenerateDataOptions) (XTrain, XTest [][]float64, yTrain, yTest []float64)
- func GenerateDataClusters(nTrain, nFeatures, nClusters int, contamination float64, rng *rand.Rand) ([][]float64, []float64)
- func GetLabelN(yTrue, yPred []float64, n int) []int
- func GetOptimalNBins(data []float64) int
- func GetOutliersInliers(X [][]float64, y []float64) (XOutliers, XInliers [][]float64)
- func InvertOrder(scores []float64) []float64
- func MaxFloat(data []float64) float64
- func Mean(data []float64) float64
- func MeanAbsoluteError(yTrue, yPred []float64) float64
- func MeanSquaredError(yTrue, yPred []float64) float64
- func MinFloat(data []float64) float64
- func Percentile(data []float64, p float64) float64
- func PrecisionAtN(yTrue []float64, yScores []float64, n int) float64
- func PrecisionScore(yTrue, yPred []int) float64
- func ROCAUCScore(yTrue []float64, yScores []float64) float64
- func RecallScore(yTrue, yPred []int) float64
- func RootMeanSquaredError(yTrue, yPred []float64) float64
- func ScoreToLabel(scores []float64, outliersFraction float64) []int
- func Standardize(X [][]float64) ([][]float64, []float64, []float64)
- func StdDev(data []float64) float64
- func Sum(data []float64) float64
- func TrainTestSplit(X [][]float64, y []float64, testSize float64, rng *rand.Rand) (XTrain, XTest [][]float64, yTrain, yTest []float64)
- func Variance(data []float64) float64
- type GenerateDataOptions
Constants ¶
This section is empty.
Variables ¶
var ErrInvalidParameter = errors.New("parameter out of valid range")
ErrInvalidParameter is returned when a parameter is out of valid range
Functions ¶
func ApplyStandardize ¶
ApplyStandardize applies standardization using pre-computed mean and std
func BalancedAccuracy ¶
BalancedAccuracy calculates the balanced accuracy score
func CheckParameter ¶
CheckParameter checks if a parameter is within the specified range
func ConfusionMatrix ¶
ConfusionMatrix calculates the confusion matrix Returns [TN, FP, FN, TP]
func GenerateData ¶
func GenerateData(opts *GenerateDataOptions) (XTrain, XTest [][]float64, yTrain, yTest []float64)
GenerateData generates synthesized data for outlier detection testing. Normal data is generated by a multivariate Gaussian distribution and outliers are generated by a uniform distribution.
func GenerateDataClusters ¶
func GenerateDataClusters(nTrain, nFeatures, nClusters int, contamination float64, rng *rand.Rand) ([][]float64, []float64)
GenerateDataClusters generates clustered data with outliers
func GetOptimalNBins ¶
GetOptimalNBins calculates the optimal number of bins using Birge-Rozenblac method
func GetOutliersInliers ¶
GetOutliersInliers separates inliers from outliers
func InvertOrder ¶
InvertOrder inverts the order of scores
func MeanAbsoluteError ¶
MeanAbsoluteError calculates the mean absolute error
func MeanSquaredError ¶
MeanSquaredError calculates the mean squared error
func Percentile ¶
Percentile calculates the p-th percentile of the data
func PrecisionAtN ¶
PrecisionAtN calculates precision at rank n
func PrecisionScore ¶
PrecisionScore calculates the precision score
func ROCAUCScore ¶
ROCAUCScore calculates the Area Under the Receiver Operating Characteristic Curve
func RecallScore ¶
RecallScore calculates the recall score
func RootMeanSquaredError ¶
RootMeanSquaredError calculates the root mean squared error
func ScoreToLabel ¶
ScoreToLabel converts raw outlier scores to binary labels
func Standardize ¶
Standardize performs Z-score normalization on data
Types ¶
type GenerateDataOptions ¶
type GenerateDataOptions struct {
NTrain int
NTest int
NFeatures int
Contamination float64
TrainOnly bool
Offset float64
RandomState *rand.Rand
}
GenerateDataOptions holds options for data generation
func DefaultGenerateDataOptions ¶
func DefaultGenerateDataOptions() *GenerateDataOptions
DefaultGenerateDataOptions returns default options for data generation
Source Files