org.wso2.carbon.ml.rest.api

Class ModelApiV11

java.lang.Object

org.wso2.carbon.ml.rest.api.MLRestAPI

org.wso2.carbon.ml.rest.api.ModelApiV11

public class ModelApiV11
extends MLRestAPI

This class is to handle REST verbs GET , POST and DELETE.

Field Summary

Fields inherited from class org.wso2.carbon.ml.rest.api.MLRestAPI

auditLog

Constructor Summary

Constructors

Constructor and Description
ModelApiV11()

Method Summary

Modifier and Type	Method and Description
javax.ws.rs.core.Response	addStorage(long modelId, org.wso2.carbon.ml.commons.domain.MLStorage storage) Create a new model storage
javax.ws.rs.core.Response	buildModel(long modelId) Build the model
javax.ws.rs.core.Response	createModel(org.wso2.carbon.ml.commons.domain.MLModelData model) Create a new Model.
javax.ws.rs.core.Response	deleteModel(long modelId) Delete a model
javax.ws.rs.core.Response	exportModel(long modelId, String mode) Download the model
javax.ws.rs.core.Response	getAllModels() Get all models
javax.ws.rs.core.Response	getModel(String modelName) Get the model data
javax.ws.rs.core.Response	getModelSummary(long modelId) Get the model summary
javax.ws.rs.core.Response	getProductRecommendations(long modelId, int userId, int noOfProducts) Get a list of recommended products for a given user using the given model.
javax.ws.rs.core.Response	getUserRecommendations(long modelId, int productId, int noOfUsers) Get a list of recommended users for a given product using the given model.
javax.ws.rs.core.Response	options()
javax.ws.rs.core.Response	predict(long modelId, List<String[]> data, double percentile, boolean skipDecoding) Make predictions using a model
javax.ws.rs.core.Response	predict(long modelId, String dataFormat, InputStream inputStream, double percentile, boolean skipDecoding) Predict using a file and return as a list of predicted values.
javax.ws.rs.core.Response	publishModel(long modelId, String mode) Publish the model to ML registry
javax.ws.rs.core.Response	streamingPredict(long modelId, String dataFormat, String columnHeader, InputStream inputStream, double percentile, boolean skipDecoding) Predict using a file and return predictions as a CSV.

Methods inherited from class org.wso2.carbon.ml.rest.api.MLRestAPI

getTenantID

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ModelApiV11

public ModelApiV11()

Method Detail

options

public javax.ws.rs.core.Response options()

createModel

public javax.ws.rs.core.Response createModel(org.wso2.carbon.ml.commons.domain.MLModelData model)

Create a new Model.

Parameters:

model - MLModelData object

Returns:

JSON of MLModelData object

addStorage

public javax.ws.rs.core.Response addStorage(long modelId,
                                            org.wso2.carbon.ml.commons.domain.MLStorage storage)

Create a new model storage

Parameters:

modelId - Unique id of the model

storage - MLStorage object

buildModel

public javax.ws.rs.core.Response buildModel(long modelId)

Build the model

Parameters:

modelId - Unique id of the model to be built.

publishModel

public javax.ws.rs.core.Response publishModel(long modelId,
                                              String mode)

Publish the model to ML registry

Parameters:

modelId - Unique id of the model to be published

Returns:

JSON of MLResponseBean containing the published location of the model

predict

public javax.ws.rs.core.Response predict(long modelId,
                                         String dataFormat,
                                         InputStream inputStream,
                                         double percentile,
                                         boolean skipDecoding)

Predict using a file and return as a list of predicted values.

Parameters:

modelId - Unique id of the model

dataFormat - Data format of the file (CSV or TSV)

inputStream - File input stream generated from the file used for predictions

percentile - a threshold value used to identified cluster boundaries

skipDecoding - whether the decoding should not be done (true or false)

Returns:

JSON array of predictions

streamingPredict

public javax.ws.rs.core.Response streamingPredict(long modelId,
                                                  String dataFormat,
                                                  String columnHeader,
                                                  InputStream inputStream,
                                                  double percentile,
                                                  boolean skipDecoding)

Predict using a file and return predictions as a CSV.

Parameters:

modelId - Unique id of the model

dataFormat - Data format of the file (CSV or TSV)

columnHeader - Whether the file contains the column header as the first row (YES or NO)

inputStream - Input stream generated from the file used for predictions

percentile - a threshold value used to identified cluster boundaries

skipDecoding - whether the decoding should not be done (true or false)

Returns:

A file as a StreamingOutput

predict

public javax.ws.rs.core.Response predict(long modelId,
                                         List<String[]> data,
                                         double percentile,
                                         boolean skipDecoding)

Make predictions using a model

Parameters:

modelId - Unique id of the model

data - List of string arrays containing the feature values used for predictions

percentile - a threshold value used to identified cluster boundaries

skipDecoding - whether the decoding should not be done (true or false)

Returns:

JSON array of predicted values

getModel

public javax.ws.rs.core.Response getModel(String modelName)

Get the model data

Parameters:

modelName - Name of the model

Returns:

JSON of MLModelData object

getAllModels

public javax.ws.rs.core.Response getAllModels()

Get all models

Returns:

JSON array of MLModelData objects

deleteModel

public javax.ws.rs.core.Response deleteModel(long modelId)

Delete a model

Parameters:

modelId - Unique id of the model

getModelSummary

public javax.ws.rs.core.Response getModelSummary(long modelId)

Get the model summary

Parameters:

modelId - Unique id of the model

Returns:

JSON of ModelSummary object

exportModel

public javax.ws.rs.core.Response exportModel(long modelId,
                                             String mode)

Download the model

Parameters:

modelId - Name of the model

Returns:

A MLModel as a StreamingOutput

getProductRecommendations

public javax.ws.rs.core.Response getProductRecommendations(long modelId,
                                                           int userId,
                                                           int noOfProducts)

Get a list of recommended products for a given user using the given model.

Parameters:

modelId - id of the recommendation model to be used.

userId - id of the user.

noOfProducts - number of recommendations required.

Returns:

an array of product recommendations.

getUserRecommendations

public javax.ws.rs.core.Response getUserRecommendations(long modelId,
                                                        int productId,
                                                        int noOfUsers)

Get a list of recommended users for a given product using the given model.

Parameters:

modelId - id of the recommendation model to be used.

productId - id of the product.

noOfUsers - number of recommendations required.

Returns:

an array of user recommendations.