R/Enums.R

In datarobot: 'DataRobot' Predictive Modeling API

namedEnumList <- function(...) {
  list <- as.list(...)
  names(list) <- make.names(list)
  list
}

#' Autopilot modes
#'
#' This is a list that contains the valid values for autopilot mode. If you wish, you can
#' specify autopilot modes using the list values, e.g. AutopilotMode$FullAuto instead of typing
#' the string "auto". This way you can benefit from autocomplete and not have to remember the valid
#' options.
#'
#' \code{FullAuto} represents running the entire autopilot. \code{Quick} runs a quicker, abridged
#' version of the autopilot that focuses on the most important models. \code{Manual} does not run
#' the autopilot and instead leaves it to the user to select the algorithms to be run.
#' \code{Comprehensive} runs all blueprints in the repository, and may be extremely slow.
#' @export
AutopilotMode <- list(
  FullAuto = "auto",
  Manual = "manual",
  Quick = "quick",
  Comprehensive = "comprehensive"
)

#' Job statuses
#'
#' This is a list that contains the valid values for job status when querying the list of jobs mode.
#' If you wish, you can specify job status modes using the list values, e.g. JobStatus$InProgress
#' instead of typing the string "inprogress". This way you can benefit from autocomplete and not
#' have to remember the valid options.
#' @export
JobStatus <- list(
  Queue = "queue",
  InProgress = "inprogress",
  Error = "error",
  Aborted = "ABORTED",
  Completed = "COMPLETED")

JobFailureStatuses <- c(JobStatus$Error, JobStatus$Aborted)


#' Job type
#'
#' This is a list that contains the valid values for job type when querying the list of jobs.
#' @export
JobType <- list(
  FeatureImpact = "featureImpact",
  Predict = "predict",
  Model = "model",
  PrimeRulesets = "primeRulesets",
  PrimeDownloadValidation = "primeDownloadValidation",
  PrimeModel = "primeModel",
  ModelExport = "modelExport",
  PredictionExplanationsInitialization = "predictionExplanationsInitialization",
  PredictionExplanations = "predictionExplanations",
  PredictionIntervals = "calculatePredictionIntervals"
)


#' Prime Language
#'
#' This is a list that contains the valid values for downloadable code programming languages.
#' @export
PrimeLanguage <- list(
  Python = "Python",
  Java = "Java")


#' PostgreSQL drivers
#'
#' This is a list that contains the valid values for PostgreSQL drivers.
#' @export
PostgreSQLdrivers <- list(
  Unicode = "PostgreSQL Unicode",
  ANSI = "PostgreSQL ANSI")


#' Blend methods
#'
#' This is a list that contains the valid values for Blend methods
#' @export
BlendMethods <- list(
  PLS = "PLS",
  GLM = "GLM",
  ENET = "ENET",
  MED = "MED",
  AVERAGE = "AVG",
  MAE = "MAE",
  MAEL1 = "MAEL1",
  RANDOM_FOREST = "RF",
  LIGHT_GBM = "LGBM",
  TENSORFLOW = "TF",
  FORECAST_DISTANCE = "FORECAST_DISTANCE",
  FORECAST_DISTANCE_ENET = "FORECAST_DISTANCE_ENET",
  FORECAST_DISTANCE_AVG = "FORECAST_DISTANCE_AVG"
 )


#' CV methods
#'
#' This is a list that contains the valid values for CV methods
#' @export
cvMethods <- list(
  RANDOM = "random",
  STRATIFIED = "stratified",
  USER = "user",
  GROUP = "group",
  DATETIME = "datetime"
 )


#' Data Partition methods
#'
#' This is a list that contains the valid values for data partitions
#' @export
DataPartition <- list(
  VALIDATION = "validation",
  CROSSVALIDATION = "crossValidation",
  HOLDOUT = "holdout"
 )


#' Source types
#'
#' This is a list that contains the valid values for source type
#' @export
SourceType <- list(
  Validation = "validation",
  Training = "training"
)


#' Target Type modes
#'
#' This is a list that contains the valid values for the Target Types
#' @export
TargetType  <- list(
    Binary = "Binary",
    Multiclass = "Multiclass",
    Regression = "Regression"
)


#' Data subset for training predictions
#'
#' This is a list that contains the valid values for the \code{dataSubset} parameter
#' found in \code{RequestTrainingPredictions}. If you wish, you can specify
#' \code{dataSubset} using the list values here.
#'
#' For \code{All}, all available data is used.
#'
#' For \code{ValidationAndHoldout}, only data outside the training set is used.
#'
#' For \code{Holdout}, only holdout data is used.
#'
#' For \code{AllBacktests}, data is used from all backtest validation folds. This requires
#' the model to have successfully scored all backtests. Backtests are available on datetime
#' partitioned projects only.
#' @export
DataSubset <- list(
  All = "all",
  ValidationAndHoldout = "validationAndHoldout",
  Holdout = "holdout",
  AllBacktests = "allBacktests")


#' Treat as exponential
#' @export
TreatAsExponential <- list(
    Always = "always",
    Never = "never",
    Auto = "auto"
)


#' Differencing method
#' @export
DifferencingMethod <- list(
    Auto = "auto",
    Simple = "simple",
    None = "none",
    Seasonal = "seasonal"
)


#' Time units
#' @export
TimeUnits <- list(
    Second = "SECOND",
    Minute = "MINUTE",
    Hour = "HOUR",
    Day = "DAY",
    Week = "WEEK",
    Month = "MONTH",
    Quarter = "QUARTER",
    Year = "YEAR")


#' Datetime trend plots resolutions
DatetimeTrendPlotsResolutions <- list(
    Milliseconds = "milliseconds",
    Seconds = "seconds",
    Minutes = "minutes",
    Hours = "hours",
    Days = "days",
    Weeks = "weeks",
    Months = "months",
    Quarters = "quarters",
    Years = "years")


#' Datetime trend plots statuses
DatetimeTrendPlotsStatuses <- list(
    Completed = "completed",
    NotCompleted = "notCompleted",
    InProgress = "inProgress",
    Errored = "errored",
    NotSupported = "notSupported",
    InsufficientData = "insufficientData")


#' Periodicity time units
#'
#' Same as time units, but kept for backwards compatibility.
#' @export
PeriodicityTimeUnits <- TimeUnits


#' Periodicity max time step
#' @export
PeriodicityMaxTimeStep <- 9223372036854775807


#' Target leakage report values
#' @export
TargetLeakageType <- list(
    SkippedDetection = "SKIPPED_DETECTION",
    False = "FALSE",
    ModerateRisk = "MODERATE_RISK",
    HighRisk = "HIGH_RISK"
)


#' Recommended model type values
#'
#' \code{MostAccurate} retrieves the most accurate model based on validation or
#' cross-validation results. In most cases, this will be a blender model.
#'
#' \code{FastAccurate} retrieves the most accurate individual model (not blender) that passes
#' set guidelines for prediction speed. If no models meet the prediction speed guideline, this
#' will not retrieve anything.
#'
#' \code{RecommendedForDeployment} retrieves the most accurate individual model. This model
#' will have undergone specific pre-preparations to be deployment ready. See
#' \code{GetModelRecommendation} for details.
#' @export
RecommendedModelType <- list(
  MostAccurate = "Most Accurate",
  FastAccurate = "Fast & Accurate",
  RecommendedForDeployment = "Recommended for Deployment"
)


#' Project stage
#' @export
ProjectStage <- list(
  AIM = "aim",
  EDA = "eda",
  EMPTY = "empty",
  MODELING = "modeling"
)


#' Sharing role
#'
#' This is a list that contains the valid values for granting access to other users (see
#' \code{Share}). If you wish, you can specify access roles using the list values, e.g.,
#' \code{SharingRole$ReadWrite} instead of typing the string "READ_WRITE". This way you can
#' benefit from autocomplete and not have to remember the valid options.
#'
#' \code{Owner} allows any action including deletion.
#'
#' \code{ReadWrite} or \code{Editor} allows modifications to the state, e.g., renaming
#' and creating data sources from a data store, but *not* deleting the entity.
#'
#' \code{ReadOnly} or \code{Consumer} - for data sources, enables creating projects and predictions;
#' for data stores, allows viewing them only.
#' @export
SharingRole <- list(
  Owner = "OWNER",
  ReadWrite = "READ_WRITE",
  User = "USER",
  Editor = "EDITOR",
  ReadOnly = "READ_ONLY",
  Consumer = "CONSUMER"
)


#' Series aggregation type
#'
#' For details, see "Calculating features across series" in the time series section of the
#' DataRobot user guide.
#' @export
SeriesAggregationType <- list(
  Average = "average",
  Total = "total"
)


#' Model replacement reason
#'
#' @export
ModelReplacementReason <- list(
  Accuracy = "ACCURACY",
  DataDrift = "DATA_DRIFT",
  Errors = "ERRORS",
  ScheduledRefresh = "SCHEDULED_REFRESH",
  ScoringSpeed = "SCORING_SPEED",
  Other = "OTHER"
)


#' Types of variable transformations
#'
#' @export
VariableTransformTypes <- list(
  Categorical = "categorical",
  CategoricalInt = "categoricalInt",
  Numeric = "numeric",
  Text = "text"
)


#' Deployment service health metrics
#'
#' Added in DataRobot API 2.18.
#'
#' For usage, see \code{GetDeploymentServiceStats}.
#' @export
DeploymentServiceHealthMetric <- list(
  TotalPredictions = "totalPredictions",
  TotalRequests = "totalRequests",
  SlowRequests = "slowRequests",
  ExecutionTime = "executionTime",
  ResponseTime = "responseTime",
  UserErrorRate = "userErrorRate",
  ServerErrorRate = "serverErrorRate",
  NumConsumers = "numConsumers",
  CacheHitRatio = "cacheHitRatio",
  MedianLoad = "medianLoad",
  PeakLoad = "peakLoad"
)

#' Segment analysis attributes
#'
#' Added in DataRobot API 2.20.
#'
#' For usage, see \code{GetDeploymentServiceStats}.
#' @export
SegmentAnalysisAttribute <- list(
  DataRobotConsumer = "DataRobot-Consumer",
  DataRobotRemoteIP = "DataRobot-Remote-IP",
  DataRobotHostName = "DataRobot-Host-Name"
)

#' Accuracy metrics for regression deployments
#'
#' Added in DataRobot API 2.18.
#' @export
RegressionDeploymentAccuracyMetric <- namedEnumList(c(
  "Gamma Deviance",
  "FVE Gamma",
  "FVE Poisson",
  "FVE Tweedie",
  "MAD",
  "MAE",
  "MAPE",
  "Poisson Deviance",
  "R Squared",
  "RMSE",
  "RMSLE",
  "Tweedie Deviance"
))

#' Accuracy metrics for classification deployments
#'
#' Added in DataRobot API 2.18.
#' @export
ClassificationDeploymentAccuracyMetric <- namedEnumList(c(
  "Accuracy",
  "AUC",
  "Balanced Accuracy",
  "FVE Binomial",
  "Gini Norm",
  "Kolmogorov-Smirnov",
  "LogLoss",
  "Rate@Top5%",
  "Rate@Top10%",
  "TPR",
  "FPR",
  "TNR",
  "PPV",
  "F1"
))

#' Accuracy metrics for multiclass deployments
#'
#' Added in DataRobot API 2.23.
#' @export
#'
MulticlassDeploymentAccuracyMetric <- namedEnumList(c(
  "LogLoss",
  "FVE Binomial",
  "FVE Multinomial"
))

#' Deployment accuracy metrics
#'
#' All possible deployment accuracy metrics. Added in DataRobot API 2.18.
#'
#' For usage, see `DeploymentAccuracy` and `DeploymentAccuracyOverTime`.
#' @export
#' @md
DeploymentAccuracyMetric <- {
  # Combine and dedupe the different metric types
  DeploymentAccuracyMetric <- c(
    RegressionDeploymentAccuracyMetric,
    ClassificationDeploymentAccuracyMetric,
    MulticlassDeploymentAccuracyMetric
  )
  DeploymentAccuracyMetric[!duplicated(DeploymentAccuracyMetric)]
}

#' Model capabilities
#'
#' For usage, see `\code{GetModelCapabilities}`.
#' @export
ModelCapability <- namedEnumList(c(
  "supportsEarlyStopping",
  "supportsImageEmbedding",
  "supportsNNVisualizations",
  "supportsImageActivationMaps",
  "supportsMonotonicConstraints",
  "supportsBlending",
  "supportsShap",
  "supportsCodeGeneration",
  "supportsModelTrainingMetrics",
  "hasParameters",
  "eligibleForPrime",
  "hasWordCloud"
))