linear

Documentation

Index

• type AlgorithmRunner
• type Config
• type Predict
  • func (p *Predict) GetPrediction(model *jamiethompsonmev1alpha1.Model, ...) (int32, error)
  • func (p *Predict) GetType() string
  • func (p *Predict) PruneHistory(model *jamiethompsonmev1alpha1.Model, ...) ([]jamiethompsonmev1alpha1.TimestampedReplicas, error)

Types

type AlgorithmRunner

type AlgorithmRunner interface {
	RunAlgorithmWithValue(algorithmPath string, value string, timeout int) (string, error)
}

Runner defines an algorithm runner, allowing algorithms to be run

type Config

type Config struct {
	StoredValues int `yaml:"storedValues"`
	LookAhead    int `yaml:"lookAhead"`
}

Config represents a linear regression prediction model configuration

type Predict

type Predict struct {
	Runner AlgorithmRunner
}

Predict provides logic for using Linear Regression to make a prediction

func (*Predict) GetPrediction

func (p *Predict) GetPrediction(model *jamiethompsonmev1alpha1.Model, replicaHistory []jamiethompsonmev1alpha1.TimestampedReplicas) (int32, error)

GetPrediction uses a linear regression to predict what the replica count should be based on historical evaluations

func (*Predict) GetType

func (p *Predict) GetType() string

GetType returns the type of the Prediction model

func (*Predict) PruneHistory

func (p *Predict) PruneHistory(model *jamiethompsonmev1alpha1.Model, replicaHistory []jamiethompsonmev1alpha1.TimestampedReplicas) ([]jamiethompsonmev1alpha1.TimestampedReplicas, error)