Class default`Abstract`

Abstract class outlining Model API operations supported by the SDK.

Hierarchy

default
- default

Index

Constructors

Methods

adjustModel batchQueryModel deleteModel describeAccuracyModel describeModel getAllModels getModel queryModel retrainModel trainModel

Constructors

constructor

new default(): default
Returns default

Methods

`Abstract` adjustModel

adjustModel(name: string, project: string, options?: AdjustOptions): Promise<void>
Partially adjusts this model with the given options.

Throws
- Something went wrong adjusting the model.
Parameters
- name: string
  
  Name of the model.
- project: string
  
  Project the model belongs to.
- Optional options: AdjustOptions
  
  Options to use when adjusting the model.
Returns Promise<void>
- Defined in src/models/modelsApiClient.ts:128

`Abstract` batchQueryModel

batchQueryModel(name: string, version: number, targetColumn: string, project: string, options: BatchQueryOptions): Promise<ModelPrediction[]>
Queries this model for a batch prediction by joining with another data source.

Returns
- All prediction results from the batch query.
Throws
- Something went wrong batch querying the model.
Parameters
- name: string
  
  Name of the model.
- version: number
- targetColumn: string
  
  Column that the model predicts.
- project: string
  
  Project the model belongs to.
- options: BatchQueryOptions
  
  Options to use when querying the model.
Returns Promise<ModelPrediction[]>
- Defined in src/models/modelsApiClient.ts:83

`Abstract` deleteModel

deleteModel(name: string, project: string): Promise<void>
Deletes this model.

Throws
- Something went wrong deleting the model.
Parameters
- name: string
  
  Name of the model.
- project: string
  
  Project the model belongs to.
Returns Promise<void>
- Defined in src/models/modelsApiClient.ts:55

`Abstract` describeAccuracyModel

describeAccuracyModel(name: string, project: string, unique_id?: string): Promise<ModelDescribeAccuracy[]>
Describes the features of this model.

Returns
- All feature descriptions of the model. Empty if the model doesn't exist.
Parameters
- name: string
  
  Name of the model.
- project: string
  
  Project the model belongs to.
- Optional unique_id: string
  
  Optional unique id to filter the accuracy by.
Returns Promise<ModelDescribeAccuracy[]>
- Defined in src/models/modelsApiClient.ts:42

`Abstract` describeModel

describeModel(name: string, project: string): Promise<ModelFeatureDescription[]>
Describes the features of this model.

Returns
- All feature descriptions of the model. Empty if the model doesn't exist.
Parameters
- name: string
  
  Name of the model.
- project: string
  
  Project the model belongs to.
Returns Promise<ModelFeatureDescription[]>
- Defined in src/models/modelsApiClient.ts:30

`Abstract` getAllModels

getAllModels(project: string): Promise<Model[]>
Gets all models associated with a project.

Returns
- All models for the given project.
Parameters
- project: string
  
  Project the models belong to.
Returns Promise<Model[]>
- Defined in src/models/modelsApiClient.ts:22

`Abstract` getModel

getModel(name: string, project: string): Promise<undefined | Model>
Gets a model by name and project.

Returns
- The matching model, or undefined if none exists.
Parameters
- name: string
  
  Name of the model.
- project: string
  
  Project the model belongs to.
Returns Promise<undefined | Model>
- Defined in src/models/modelsApiClient.ts:15

`Abstract` queryModel

queryModel(name: string, version: number, targetColumn: string, project: string, options: QueryOptions): Promise<ModelPrediction>
Queries this model for a single prediction. For batch predictions, use batchQuery.

Returns
- The prediction result.
Throws
- Something went wrong querying the model.
Parameters
- name: string
  
  Name of the model.
- version: number
- targetColumn: string
  
  Column that the model predicts.
- project: string
  
  Project the model belongs to.
- options: QueryOptions
  
  Options to use when querying the model.
Returns Promise<ModelPrediction>
- Defined in src/models/modelsApiClient.ts:66

`Abstract` retrainModel

retrainModel(name: string, targetColumn: string, project: string, options?: TrainingOptions): Promise<void>
Rerains this model with the given options.

Throws
- Something went wrong retraining the model.
Parameters
- name: string
  
  Name of the model.
- targetColumn: string
  
  Column for the model to predict.
- project: string
  
  Project the model belongs to.
- Optional options: TrainingOptions
  
  Options to use when retraining the model.
Returns Promise<void>
- Defined in src/models/modelsApiClient.ts:114

`Abstract` trainModel

trainModel(name: string, targetColumn: string, project: string, options: TrainingOptions): Promise<Model>
Trains this model with the given options.

Throws
- Something went wrong training the model.
Parameters
- name: string
  
  Name of the model.
- targetColumn: string
  
  Column for the model to predict.
- project: string
  
  Project the model belongs to.
- options: TrainingOptions
  
  Options to use when training the model.
Returns Promise<Model>
- Defined in src/models/modelsApiClient.ts:99

Class defaultAbstract

Hierarchy

Index

Constructors

Methods

Constructors

constructor

Returns default

Methods

Abstract adjustModel

Throws

Parameters

name: string

project: string

Optional options: AdjustOptions

Returns Promise<void>

Abstract batchQueryModel

Returns

Throws

Parameters

name: string

version: number

targetColumn: string

project: string

options: BatchQueryOptions

Returns Promise<ModelPrediction[]>

Abstract deleteModel

Throws

Parameters

name: string

project: string

Returns Promise<void>

Abstract describeAccuracyModel

Returns

Parameters

name: string

project: string

Optional unique_id: string

Returns Promise<ModelDescribeAccuracy[]>

Abstract describeModel

Returns

Parameters

name: string

project: string

Returns Promise<ModelFeatureDescription[]>

Abstract getAllModels

Returns

Parameters

project: string

Returns Promise<Model[]>

Abstract getModel

Returns

Parameters

name: string

project: string

Returns Promise<undefined | Model>

Abstract queryModel

Returns

Throws

Parameters

name: string

version: number

targetColumn: string

project: string

options: QueryOptions

Returns Promise<ModelPrediction>

Abstract retrainModel

Throws

Parameters

name: string

targetColumn: string

project: string

Optional options: TrainingOptions

Returns Promise<void>

Abstract trainModel

Throws

Parameters

name: string

targetColumn: string

project: string

Class default`Abstract`

`Abstract` adjustModel

`Optional` options: AdjustOptions

`Abstract` batchQueryModel

`Abstract` deleteModel

`Abstract` describeAccuracyModel

`Optional` unique_id: string

`Abstract` describeModel

`Abstract` getAllModels

`Abstract` getModel

`Abstract` queryModel

`Abstract` retrainModel

`Optional` options: TrainingOptions

`Abstract` trainModel