Class ModelX

Hierarchy

  • ModelX

Implements

  • ModelContext

Constructors

Properties

Methods

Constructors

  • Parameters

    • configuration: ModelConfiguration
    • options: ModelOptions = {}

    Returns ModelX

Properties

DataSet: DataSet
Model: any
auto_assign_features?: boolean
config: ModelConfiguration
cross_validation_options: CrossValidationOptions
debug: boolean
dependent_variables?: string[]
dimension?: Dimensions
emptyObject: Datum
entity?: Entity
forecastDates: Date[]
getTimeseriesDimension: timeseriesCalculation
independent_variables?: string[]
input_independent_features?: AutoFeature[]
max_evaluation_outputs: number
mockEncodedData: Data
original_data_test: Data
original_data_train: Data
output_dependent_features?: AutoFeature[]
prediction_inputs?: Data
prediction_inputs_next_value_function?: ForecastPredictionInputNextValueFunction
prediction_inputs_next_value_functions: GeneratedFunctionDefinitionsList
prediction_options?: PredictModelConfig
prediction_timeseries_date_feature: string
prediction_timeseries_date_format?: string
prediction_timeseries_dimension_feature: string
prediction_timeseries_end_date?: string | Date
prediction_timeseries_start_date?: string | Date
prediction_timeseries_time_zone: string
preprocessing_feature_column_options: DataSetTransform
removedFilterdtrainingData: Data
retrain_forecast_model_with_predictions?: boolean
scikit: any
status: ModelStatus
testDataSet: DataSet
tf: any
trainDataSet: DataSet
trainingData: Data
training_data_filter_function?: DataFilterFunction
training_data_filter_function_body?: string
training_feature_column_options: DataSetTransform
training_model_loss?: number
training_options: TensorScriptOptions
training_progress_callback: TrainingProgressCallback
training_size_values?: number
use_empty_objects: boolean
use_mock_encoded_data: boolean
use_next_value_functions_for_training_data: boolean
use_preprocessing_on_trainning_data: boolean
validate_training_data: boolean
x_indep_matrix_test: Matrix
x_indep_matrix_train: Matrix
x_independent_features: string[]
x_raw_independent_features: string[]
y_dep_matrix_test: Matrix
y_dep_matrix_train: Matrix
y_dependent_labels: string[]
y_raw_dependent_labels: string[]
dimension: Dimensions
prediction_timeseries_date_feature: string
prediction_timeseries_date_format: string
prediction_timeseries_dimension_feature: string

Methods

  • Parameters

    • __namedParameters: { use_mock_dates: boolean } = {}
      • use_mock_dates: boolean

    Returns void

  • Parameters

    • options: { retrain?: boolean } = {}
      • Optional retrain?: boolean

    Returns Promise<boolean>

  • Parameters

    • options: EvaluationAccuracyOptions = {}

    Returns { accuracy: any; actuals: any; estimates: any; labels: any; matrix: any }

    • accuracy: any
    • actuals: any
    • estimates: any
    • labels: any
    • matrix: any
  • Parameters

    • options: EvaluateModelOptions = {}

    Returns Promise<EvaluateClassificationModel | EvaluateRegressionModel>

  • Parameters

    • options: EvaluationAccuracyOptions = {}

    Returns RegressionEvaluation

  • Returns { test: any; train: any }

    • test: any
    • train: any
  • Parameters

    • options: GetDataSetProperties = {}

    Returns Promise<void>

  • Parameters

    • options: {} = {}

      Returns Date[]

    • Parameters

      • options: ValidateTimeseriesDataOptions = {}

      Returns { forecastDates: any[]; raw_prediction_inputs: {}[] }

      • forecastDates: any[]
      • raw_prediction_inputs: {}[]
    • Parameters

      • options: { getPredictionInputPromise?: GetPredicitonData } = {}
        • Optional getPredictionInputPromise?: GetPredicitonData

      Returns Promise<Data>

    • Parameters

      • options: { getDataPromise?: GetPredicitonData; retrain?: boolean; trainingData?: Data } = {}
        • Optional getDataPromise?: GetPredicitonData
        • Optional retrain?: boolean
        • Optional trainingData?: Data

      Returns Promise<void>

    • Parameters

      • options: Data | PredictModelOptions = {}

      Returns Promise<any>

    • Parameters

      • __namedParameters: { use_mock_dates: boolean } = {}
        • use_mock_dates: boolean

      Returns void

    • Parameters

      • options: retrainTimeseriesModel = {}

      Returns Promise<ModelX>

    • Parameters

      • options: ValidateTimeseriesDataOptions = {}

      Returns Promise<any[]>

    • Parameters

      • options: ModelTrainningOptions = {}

      Returns Promise<ModelX>

    • Parameters

      • options: ValidateTimeseriesDataOptions = {}

      Returns Promise<{ datasetDates: any; dimension: Dimensions; forecastDateFirstDataSetDateIndex: any; forecastDates: Date[]; lastOriginalForecastDate: Date; raw_prediction_inputs: Data }>

    • Parameters

      • __namedParameters: { cross_validate_training_data?: boolean; inputMatrix?: Matrix } = ...
        • Optional cross_validate_training_data?: boolean
        • Optional inputMatrix?: Matrix

      Returns boolean

    • Attempts to automatically figure out the time dimension of each date feature (hourly, daily, etc) and the format of the date property (e.g. JS Date Object, or ISO String, etc) from the dataset data

      Parameters

      • options: { DataSetData?: Datum; dimension?: Dimensions; timeseries_date_feature?: string; timeseries_date_format?: string } = {}
        • Optional DataSetData?: Datum
        • Optional dimension?: Dimensions
        • Optional timeseries_date_feature?: string
        • Optional timeseries_date_format?: string

      Returns TimeseriesDimension