Scoring
Table of Contents
Constants
- EVENT_ACTIVITY = 'activity'
- EVENT_ENTITY_UPDATE = 'update'
- EVENT_INITIAL_PREDICTION = 'initial'
- PREDICTION_BATCH = 'batch'
- PREDICTION_IMMEDIATE = 'immediate'
- PREDICTION_REAL_TIME = 'realtime'
- RETRAIN_PERIOD = 90
Methods
- canStartTraining() : Result
- Checks if scoring model is suitable to start training.
- deleteMlModel() : Result
- deletePrediction() : bool
- Deletes prediction with the given id.
- getAvailableModelNames() : mixed
- getCurrentPrediction() : array<string|int, mixed>|false
- Returns current prediction record or false if prediction is not found.
- getLastTraining() : array<string|int, mixed>|false
- Return current training fields for the specified model.
- getLicenseInfoText() : string
- getLicenseInfoTitle() : string
- getMinimalClassSize() : mixed
- getMinimalTrainingSetSize() : mixed
- getModelByName() : Base
- getModelClasses() : array<string|int, mixed>
- Returns available classes to work with scoring models.
- getPredictionUpdatePullTag() : string
- getScoringAvailableDate() : int
- getScoringModel() : Base|null
- hasAccess() : bool
- isEnabled() : bool
- Returns true if scoring is enabled for this portal by the tariffs.
- isMlAvailable() : bool
- Returns true if machine learning is installed for this instance.
- isScoringAvailable() : bool
- isTrainingUsed() : bool
- onActivityDelete() : void
- Removes references to this activity
- onEntityDelete() : void
- Deletes prediction history records, associated with the entity.
- onMlModelStateChange() : void
- queuePredictionUpdate() : int|false
- Return id of the scheduled request.
- replaceAssociatedEntity() : void
- Replaces entity type and id in history records.
- sendPredictionUpdatePullEvent() : void
- setScoringAvailableDate() : void
- startModelTraining() : Result
- Starts training of the model, if all of the pre-requirements are met, such as: - ml model should not exists, trainer will create it later - last training should be in state finished
- tryCreateFirstPrediction() : bool
- Tries to create first prediction for the given entity. Preconditions: - ml module should be installed - model for this entity type should be in ready state - this entity should not have another predictions
- updatePrediction() : Result
Constants
EVENT_ACTIVITY
public
mixed
EVENT_ACTIVITY
= 'activity'
EVENT_ENTITY_UPDATE
public
mixed
EVENT_ENTITY_UPDATE
= 'update'
EVENT_INITIAL_PREDICTION
public
mixed
EVENT_INITIAL_PREDICTION
= 'initial'
PREDICTION_BATCH
public
mixed
PREDICTION_BATCH
= 'batch'
PREDICTION_IMMEDIATE
public
mixed
PREDICTION_IMMEDIATE
= 'immediate'
PREDICTION_REAL_TIME
public
mixed
PREDICTION_REAL_TIME
= 'realtime'
RETRAIN_PERIOD
public
mixed
RETRAIN_PERIOD
= 90
Methods
canStartTraining()
Checks if scoring model is suitable to start training.
public
static canStartTraining(Base $model[, bool $useCache = false ]) : Result
Parameters
- $model : Base
-
Scoring model.
- $useCache : bool = false
Return values
ResultdeleteMlModel()
public
static deleteMlModel(Base $model) : Result
Parameters
- $model : Base
Return values
ResultdeletePrediction()
Deletes prediction with the given id.
public
static deletePrediction(int $historyId) : bool
Parameters
- $historyId : int
Return values
boolgetAvailableModelNames()
public
static getAvailableModelNames() : mixed
getCurrentPrediction()
Returns current prediction record or false if prediction is not found.
public
static getCurrentPrediction(int $entityTypeId, int $entityId) : array<string|int, mixed>|false
Parameters
- $entityTypeId : int
-
Entity type id.
- $entityId : int
-
Id of the entity.
Return values
array<string|int, mixed>|falsegetLastTraining()
Return current training fields for the specified model.
public
static getLastTraining(Base $model) : array<string|int, mixed>|false
Parameters
- $model : Base
Return values
array<string|int, mixed>|falsegetLicenseInfoText()
public
static getLicenseInfoText() : string
Return values
stringgetLicenseInfoTitle()
public
static getLicenseInfoTitle() : string
Return values
stringgetMinimalClassSize()
public
static getMinimalClassSize() : mixed
getMinimalTrainingSetSize()
public
static getMinimalTrainingSetSize() : mixed
getModelByName()
public
static getModelByName(mixed $modelName) : Base
Parameters
- $modelName : mixed
Return values
BasegetModelClasses()
Returns available classes to work with scoring models.
public
static getModelClasses() : array<string|int, mixed>
Return values
array<string|int, mixed>getPredictionUpdatePullTag()
public
static getPredictionUpdatePullTag(int $entityTypeId, int $entityId) : string
Parameters
- $entityTypeId : int
- $entityId : int
Return values
stringgetScoringAvailableDate()
public
static getScoringAvailableDate() : int
Return values
intgetScoringModel()
public
static getScoringModel(int $entityTypeId, int $entityId) : Base|null
Parameters
- $entityTypeId : int
- $entityId : int
Return values
Base|nullhasAccess()
public
static hasAccess(string $modelName[, int $userId = 0 ]) : bool
Parameters
- $modelName : string
- $userId : int = 0
Return values
boolisEnabled()
Returns true if scoring is enabled for this portal by the tariffs.
public
static isEnabled() : bool
Return values
boolisMlAvailable()
Returns true if machine learning is installed for this instance.
public
static isMlAvailable() : bool
Return values
boolisScoringAvailable()
public
static isScoringAvailable() : bool
Return values
boolisTrainingUsed()
public
static isTrainingUsed() : bool
Return values
boolonActivityDelete()
Removes references to this activity
public
static onActivityDelete(int $activityId) : void
Parameters
- $activityId : int
onEntityDelete()
Deletes prediction history records, associated with the entity.
public
static onEntityDelete(int $entityTypeId, int $entityId) : void
Parameters
- $entityTypeId : int
-
Entity type.
- $entityId : int
-
Entity id.
onMlModelStateChange()
public
static onMlModelStateChange(Event $event) : void
Parameters
- $event : Event
queuePredictionUpdate()
Return id of the scheduled request.
public
static queuePredictionUpdate(int $entityTypeId, int $entityId[, array<string|int, mixed> $additionalParameters = [] ]) : int|false
Parameters
- $entityTypeId : int
- $entityId : int
- $additionalParameters : array<string|int, mixed> = []
-
- EVENT_TYPE
- ASSOCIATED_ACTIVITY_ID
Return values
int|falsereplaceAssociatedEntity()
Replaces entity type and id in history records.
public
static replaceAssociatedEntity(int $entityTypeId, int $entityId, mixed $newEntityTypeId, mixed $newEntityId) : void
Parameters
- $entityTypeId : int
-
Old entity type.
- $entityId : int
-
Old entity id.
- $newEntityTypeId : mixed
- $newEntityId : mixed
sendPredictionUpdatePullEvent()
public
static sendPredictionUpdatePullEvent(mixed $entityTypeId, mixed $entityId, mixed $predictionRecord) : void
Parameters
- $entityTypeId : mixed
- $entityId : mixed
- $predictionRecord : mixed
setScoringAvailableDate()
public
static setScoringAvailableDate(int $value) : void
Parameters
- $value : int
startModelTraining()
Starts training of the model, if all of the pre-requirements are met, such as: - ml model should not exists, trainer will create it later - last training should be in state finished
public
static startModelTraining(Base $model) : Result
Parameters
- $model : Base
-
Scoring model to train.
Return values
ResulttryCreateFirstPrediction()
Tries to create first prediction for the given entity. Preconditions: - ml module should be installed - model for this entity type should be in ready state - this entity should not have another predictions
public
static tryCreateFirstPrediction(int $entityTypeId, int $entityId[, bool $isImmediate = false ]) : bool
Parameters
- $entityTypeId : int
-
Type of the entity.
- $entityId : int
-
Id of the entity.
- $isImmediate : bool = false
-
Should prediction request be executed immediately.
Return values
boolupdatePrediction()
public
static updatePrediction(mixed $entityTypeId, mixed $entityId[, array<string|int, mixed> $parameters = [] ]) : Result
Parameters
- $entityTypeId : mixed
- $entityId : mixed
- $parameters : array<string|int, mixed> = []
-
- EVENT_TYPE string
- ASSOCIATED_ACTIVITY_ID int