prediction/v1.3/prediction-gen.go - google-api-go-client - Git at Google

 // Package prediction provides access to the Prediction API.
 //
 // See https://developers.google.com/prediction/docs/developer-guide
 //
 // Usage example:
 //
 //   import "google.golang.org/api/prediction/v1.3"
 //   ...
 //   predictionService, err := prediction.New(oauthHttpClient)
 package prediction

 import (
 	"bytes"
 	"encoding/json"
 	"errors"
 	"fmt"
 	"golang.org/x/net/context"
 	"google.golang.org/api/googleapi"
 	"io"
 	"net/http"
 	"net/url"
 	"strconv"
 	"strings"
 )

 // Always reference these packages, just in case the auto-generated code
 // below doesn't.
 var _ = bytes.NewBuffer
 var _ = strconv.Itoa
 var _ = fmt.Sprintf
 var _ = json.NewDecoder
 var _ = io.Copy
 var _ = url.Parse
 var _ = googleapi.Version
 var _ = errors.New
 var _ = strings.Replace
 var _ = context.Background

 const apiId = "prediction:v1.3"
 const apiName = "prediction"
 const apiVersion = "v1.3"
 const basePath = "https://www.googleapis.com/prediction/v1.3/"

 // OAuth2 scopes used by this API.
 const (
 	// Manage your data and permissions in Google Cloud Storage
 	DevstorageFull_controlScope = "https://www.googleapis.com/auth/devstorage.full_control"

 	// View your data in Google Cloud Storage
 	DevstorageRead_onlyScope = "https://www.googleapis.com/auth/devstorage.read_only"

 	// Manage your data in Google Cloud Storage
 	DevstorageRead_writeScope = "https://www.googleapis.com/auth/devstorage.read_write"

 	// Manage your data in the Google Prediction API
 	PredictionScope = "https://www.googleapis.com/auth/prediction"
 )

 func New(client *http.Client) (*Service, error) {
 	if client == nil {
 		return nil, errors.New("client is nil")
 	}
 	s := &Service{client: client, BasePath: basePath}
 	s.Hostedmodels = NewHostedmodelsService(s)
 	s.Training = NewTrainingService(s)
 	return s, nil
 }

 type Service struct {
 	client   *http.Client
 	BasePath string // API endpoint base URL

 	Hostedmodels *HostedmodelsService

 	Training *TrainingService
 }

 func NewHostedmodelsService(s *Service) *HostedmodelsService {
 	rs := &HostedmodelsService{s: s}
 	return rs
 }

 type HostedmodelsService struct {
 	s *Service
 }

 func NewTrainingService(s *Service) *TrainingService {
 	rs := &TrainingService{s: s}
 	return rs
 }

 type TrainingService struct {
 	s *Service
 }

 type Input struct {
 	// Input: Input to the model for a prediction
 	Input *InputInput `json:"input,omitempty"`
 }

 type InputInput struct {
 	// CsvInstance: A list of input features, these can be strings or
 	// doubles.
 	CsvInstance []interface{} `json:"csvInstance,omitempty"`
 }

 type Output struct {
 	// Id: The unique name for the predictive model.
 	Id string `json:"id,omitempty"`

 	// Kind: What kind of resource this is.
 	Kind string `json:"kind,omitempty"`

 	// OutputLabel: The most likely class [Categorical models only].
 	OutputLabel string `json:"outputLabel,omitempty"`

 	// OutputMulti: A list of classes with their estimated probabilities
 	// [Categorical models only].
 	OutputMulti []*OutputOutputMulti `json:"outputMulti,omitempty"`

 	// OutputValue: The estimated regression value [Regression models only].
 	OutputValue float64 `json:"outputValue,omitempty"`

 	// SelfLink: A URL to re-request this resource.
 	SelfLink string `json:"selfLink,omitempty"`
 }

 type OutputOutputMulti struct {
 	// Label: The class label.
 	Label string `json:"label,omitempty"`

 	// Score: The probability of the class.
 	Score float64 `json:"score,omitempty"`
 }

 type Training struct {
 	// Id: The unique name for the predictive model.
 	Id string `json:"id,omitempty"`

 	// Kind: What kind of resource this is.
 	Kind string `json:"kind,omitempty"`

 	// ModelInfo: Model metadata.
 	ModelInfo *TrainingModelInfo `json:"modelInfo,omitempty"`

 	// SelfLink: A URL to re-request this resource.
 	SelfLink string `json:"selfLink,omitempty"`

 	// TrainingStatus: The current status of the training job. This can be
 	// one of following: RUNNING; DONE; ERROR; ERROR: TRAINING JOB NOT FOUND
 	TrainingStatus string `json:"trainingStatus,omitempty"`

 	// Utility: A class weighting function, which allows the importance
 	// weights for classes to be specified [Categorical models only].
 	Utility []*TrainingUtility `json:"utility,omitempty"`
 }

 type TrainingModelInfo struct {
 	// ClassWeightedAccuracy: Estimated accuracy of model taking utility
 	// weights into account [Categorical models only].
 	ClassWeightedAccuracy float64 `json:"classWeightedAccuracy,omitempty"`

 	// ClassificationAccuracy: A number between 0.0 and 1.0, where 1.0 is
 	// 100% accurate. This is an estimate, based on the amount and quality
 	// of the training data, of the estimated prediction accuracy. You can
 	// use this is a guide to decide whether the results are accurate enough
 	// for your needs. This estimate will be more reliable if your real
 	// input data is similar to your training data [Categorical models
 	// only].
 	ClassificationAccuracy float64 `json:"classificationAccuracy,omitempty"`

 	// ConfusionMatrix: An output confusion matrix. This shows an estimate
 	// for how this model will do in predictions. This is first indexed by
 	// the true class label. For each true class label, this provides a pair
 	// {predicted_label, count}, where count is the estimated number of
 	// times the model will predict the predicted label given the true
 	// label. Will not output if more then 100 classes [Categorical models
 	// only].
 	ConfusionMatrix *TrainingModelInfoConfusionMatrix `json:"confusionMatrix,omitempty"`

 	// ConfusionMatrixRowTotals: A list of the confusion matrix row totals
 	ConfusionMatrixRowTotals *TrainingModelInfoConfusionMatrixRowTotals `json:"confusionMatrixRowTotals,omitempty"`

 	// MeanSquaredError: An estimated mean squared error. The can be used to
 	// measure the quality of the predicted model [Regression models only].
 	MeanSquaredError float64 `json:"meanSquaredError,omitempty"`

 	// ModelType: Type of predictive model (CLASSIFICATION or REGRESSION)
 	ModelType string `json:"modelType,omitempty"`

 	// NumberClasses: Number of classes in the trained model [Categorical
 	// models only].
 	NumberClasses int64 `json:"numberClasses,omitempty,string"`

 	// NumberInstances: Number of valid data instances used in the trained
 	// model.
 	NumberInstances int64 `json:"numberInstances,omitempty,string"`
 }

 type TrainingModelInfoConfusionMatrix struct {
 }

 type TrainingModelInfoConfusionMatrixRowTotals struct {
 }

 type TrainingUtility struct {
 }

 type Update struct {
 	// ClassLabel: The true class label of this instance
 	ClassLabel string `json:"classLabel,omitempty"`

 	// CsvInstance: The input features for this instance
 	CsvInstance []interface{} `json:"csvInstance,omitempty"`
 }

 // method id "prediction.hostedmodels.predict":

 type HostedmodelsPredictCall struct {
 	s               *Service
 	hostedModelName string
 	input           *Input
 	opt_            map[string]interface{}
 }

 // Predict: Submit input and request an output against a hosted model
 func (r *HostedmodelsService) Predict(hostedModelName string, input *Input) *HostedmodelsPredictCall {
 	c := &HostedmodelsPredictCall{s: r.s, opt_: make(map[string]interface{})}
 	c.hostedModelName = hostedModelName
 	c.input = input
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *HostedmodelsPredictCall) Fields(s ...googleapi.Field) *HostedmodelsPredictCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *HostedmodelsPredictCall) Do() (*Output, error) {
 	var body io.Reader = nil
 	body, err := googleapi.WithoutDataWrapper.JSONReader(c.input)
 	if err != nil {
 		return nil, err
 	}
 	ctype := "application/json"
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "hostedmodels/{hostedModelName}/predict")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("POST", urls, body)
 	googleapi.Expand(req.URL, map[string]string{
 		"hostedModelName": c.hostedModelName,
 	})
 	req.Header.Set("Content-Type", ctype)
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return nil, err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return nil, err
 	}
 	var ret *Output
 	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
 		return nil, err
 	}
 	return ret, nil
 	// {
 	//   "description": "Submit input and request an output against a hosted model",
 	//   "httpMethod": "POST",
 	//   "id": "prediction.hostedmodels.predict",
 	//   "parameterOrder": [
 	//     "hostedModelName"
 	//   ],
 	//   "parameters": {
 	//     "hostedModelName": {
 	//       "description": "The name of a hosted model",
 	//       "location": "path",
 	//       "required": true,
 	//       "type": "string"
 	//     }
 	//   },
 	//   "path": "hostedmodels/{hostedModelName}/predict",
 	//   "request": {
 	//     "$ref": "Input"
 	//   },
 	//   "response": {
 	//     "$ref": "Output"
 	//   },
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }

 // method id "prediction.training.delete":

 type TrainingDeleteCall struct {
 	s    *Service
 	data string
 	opt_ map[string]interface{}
 }

 // Delete: Delete a trained model
 func (r *TrainingService) Delete(data string) *TrainingDeleteCall {
 	c := &TrainingDeleteCall{s: r.s, opt_: make(map[string]interface{})}
 	c.data = data
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *TrainingDeleteCall) Fields(s ...googleapi.Field) *TrainingDeleteCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *TrainingDeleteCall) Do() error {
 	var body io.Reader = nil
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("DELETE", urls, body)
 	googleapi.Expand(req.URL, map[string]string{
 		"data": c.data,
 	})
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return err
 	}
 	return nil
 	// {
 	//   "description": "Delete a trained model",
 	//   "httpMethod": "DELETE",
 	//   "id": "prediction.training.delete",
 	//   "parameterOrder": [
 	//     "data"
 	//   ],
 	//   "parameters": {
 	//     "data": {
 	//       "description": "mybucket/mydata resource in Google Storage",
 	//       "location": "path",
 	//       "required": true,
 	//       "type": "string"
 	//     }
 	//   },
 	//   "path": "training/{data}",
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }

 // method id "prediction.training.get":

 type TrainingGetCall struct {
 	s    *Service
 	data string
 	opt_ map[string]interface{}
 }

 // Get: Check training status of your model
 func (r *TrainingService) Get(data string) *TrainingGetCall {
 	c := &TrainingGetCall{s: r.s, opt_: make(map[string]interface{})}
 	c.data = data
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *TrainingGetCall) Fields(s ...googleapi.Field) *TrainingGetCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *TrainingGetCall) Do() (*Training, error) {
 	var body io.Reader = nil
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("GET", urls, body)
 	googleapi.Expand(req.URL, map[string]string{
 		"data": c.data,
 	})
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return nil, err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return nil, err
 	}
 	var ret *Training
 	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
 		return nil, err
 	}
 	return ret, nil
 	// {
 	//   "description": "Check training status of your model",
 	//   "httpMethod": "GET",
 	//   "id": "prediction.training.get",
 	//   "parameterOrder": [
 	//     "data"
 	//   ],
 	//   "parameters": {
 	//     "data": {
 	//       "description": "mybucket/mydata resource in Google Storage",
 	//       "location": "path",
 	//       "required": true,
 	//       "type": "string"
 	//     }
 	//   },
 	//   "path": "training/{data}",
 	//   "response": {
 	//     "$ref": "Training"
 	//   },
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }

 // method id "prediction.training.insert":

 type TrainingInsertCall struct {
 	s        *Service
 	training *Training
 	opt_     map[string]interface{}
 }

 // Insert: Begin training your model
 func (r *TrainingService) Insert(training *Training) *TrainingInsertCall {
 	c := &TrainingInsertCall{s: r.s, opt_: make(map[string]interface{})}
 	c.training = training
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *TrainingInsertCall) Fields(s ...googleapi.Field) *TrainingInsertCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *TrainingInsertCall) Do() (*Training, error) {
 	var body io.Reader = nil
 	body, err := googleapi.WithoutDataWrapper.JSONReader(c.training)
 	if err != nil {
 		return nil, err
 	}
 	ctype := "application/json"
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "training")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("POST", urls, body)
 	googleapi.SetOpaque(req.URL)
 	req.Header.Set("Content-Type", ctype)
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return nil, err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return nil, err
 	}
 	var ret *Training
 	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
 		return nil, err
 	}
 	return ret, nil
 	// {
 	//   "description": "Begin training your model",
 	//   "httpMethod": "POST",
 	//   "id": "prediction.training.insert",
 	//   "path": "training",
 	//   "request": {
 	//     "$ref": "Training"
 	//   },
 	//   "response": {
 	//     "$ref": "Training"
 	//   },
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/devstorage.full_control",
 	//     "https://www.googleapis.com/auth/devstorage.read_only",
 	//     "https://www.googleapis.com/auth/devstorage.read_write",
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }

 // method id "prediction.training.predict":

 type TrainingPredictCall struct {
 	s     *Service
 	data  string
 	input *Input
 	opt_  map[string]interface{}
 }

 // Predict: Submit data and request a prediction
 func (r *TrainingService) Predict(data string, input *Input) *TrainingPredictCall {
 	c := &TrainingPredictCall{s: r.s, opt_: make(map[string]interface{})}
 	c.data = data
 	c.input = input
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *TrainingPredictCall) Fields(s ...googleapi.Field) *TrainingPredictCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *TrainingPredictCall) Do() (*Output, error) {
 	var body io.Reader = nil
 	body, err := googleapi.WithoutDataWrapper.JSONReader(c.input)
 	if err != nil {
 		return nil, err
 	}
 	ctype := "application/json"
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}/predict")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("POST", urls, body)
 	googleapi.Expand(req.URL, map[string]string{
 		"data": c.data,
 	})
 	req.Header.Set("Content-Type", ctype)
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return nil, err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return nil, err
 	}
 	var ret *Output
 	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
 		return nil, err
 	}
 	return ret, nil
 	// {
 	//   "description": "Submit data and request a prediction",
 	//   "httpMethod": "POST",
 	//   "id": "prediction.training.predict",
 	//   "parameterOrder": [
 	//     "data"
 	//   ],
 	//   "parameters": {
 	//     "data": {
 	//       "description": "mybucket/mydata resource in Google Storage",
 	//       "location": "path",
 	//       "required": true,
 	//       "type": "string"
 	//     }
 	//   },
 	//   "path": "training/{data}/predict",
 	//   "request": {
 	//     "$ref": "Input"
 	//   },
 	//   "response": {
 	//     "$ref": "Output"
 	//   },
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }

 // method id "prediction.training.update":

 type TrainingUpdateCall struct {
 	s      *Service
 	data   string
 	update *Update
 	opt_   map[string]interface{}
 }

 // Update: Add new data to a trained model
 func (r *TrainingService) Update(data string, update *Update) *TrainingUpdateCall {
 	c := &TrainingUpdateCall{s: r.s, opt_: make(map[string]interface{})}
 	c.data = data
 	c.update = update
 	return c
 }

 // Fields allows partial responses to be retrieved.
 // See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
 // for more information.
 func (c *TrainingUpdateCall) Fields(s ...googleapi.Field) *TrainingUpdateCall {
 	c.opt_["fields"] = googleapi.CombineFields(s)
 	return c
 }

 func (c *TrainingUpdateCall) Do() (*Training, error) {
 	var body io.Reader = nil
 	body, err := googleapi.WithoutDataWrapper.JSONReader(c.update)
 	if err != nil {
 		return nil, err
 	}
 	ctype := "application/json"
 	params := make(url.Values)
 	params.Set("alt", "json")
 	if v, ok := c.opt_["fields"]; ok {
 		params.Set("fields", fmt.Sprintf("%v", v))
 	}
 	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
 	urls += "?" + params.Encode()
 	req, _ := http.NewRequest("PUT", urls, body)
 	googleapi.Expand(req.URL, map[string]string{
 		"data": c.data,
 	})
 	req.Header.Set("Content-Type", ctype)
 	req.Header.Set("User-Agent", "google-api-go-client/0.5")
 	res, err := c.s.client.Do(req)
 	if err != nil {
 		return nil, err
 	}
 	defer googleapi.CloseBody(res)
 	if err := googleapi.CheckResponse(res); err != nil {
 		return nil, err
 	}
 	var ret *Training
 	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
 		return nil, err
 	}
 	return ret, nil
 	// {
 	//   "description": "Add new data to a trained model",
 	//   "httpMethod": "PUT",
 	//   "id": "prediction.training.update",
 	//   "parameterOrder": [
 	//     "data"
 	//   ],
 	//   "parameters": {
 	//     "data": {
 	//       "description": "mybucket/mydata resource in Google Storage",
 	//       "location": "path",
 	//       "required": true,
 	//       "type": "string"
 	//     }
 	//   },
 	//   "path": "training/{data}",
 	//   "request": {
 	//     "$ref": "Update"
 	//   },
 	//   "response": {
 	//     "$ref": "Training"
 	//   },
 	//   "scopes": [
 	//     "https://www.googleapis.com/auth/prediction"
 	//   ]
 	// }

 }
	// Package prediction provides access to the Prediction API.
	//
	// See https://developers.google.com/prediction/docs/developer-guide
	//
	// Usage example:
	//
	// import "google.golang.org/api/prediction/v1.3"
	// ...
	// predictionService, err := prediction.New(oauthHttpClient)
	package prediction

	import (
	"bytes"
	"encoding/json"
	"errors"
	"fmt"
	"golang.org/x/net/context"
	"google.golang.org/api/googleapi"
	"io"
	"net/http"
	"net/url"
	"strconv"
	"strings"
	)

	// Always reference these packages, just in case the auto-generated code
	// below doesn't.
	var _ = bytes.NewBuffer
	var _ = strconv.Itoa
	var _ = fmt.Sprintf
	var _ = json.NewDecoder
	var _ = io.Copy
	var _ = url.Parse
	var _ = googleapi.Version
	var _ = errors.New
	var _ = strings.Replace
	var _ = context.Background

	const apiId = "prediction:v1.3"
	const apiName = "prediction"
	const apiVersion = "v1.3"
	const basePath = "https://www.googleapis.com/prediction/v1.3/"

	// OAuth2 scopes used by this API.
	const (
	// Manage your data and permissions in Google Cloud Storage
	DevstorageFull_controlScope = "https://www.googleapis.com/auth/devstorage.full_control"

	// View your data in Google Cloud Storage
	DevstorageRead_onlyScope = "https://www.googleapis.com/auth/devstorage.read_only"

	// Manage your data in Google Cloud Storage
	DevstorageRead_writeScope = "https://www.googleapis.com/auth/devstorage.read_write"

	// Manage your data in the Google Prediction API
	PredictionScope = "https://www.googleapis.com/auth/prediction"
	)

	func New(client http.Client) (Service, error) {
	if client == nil {
	return nil, errors.New("client is nil")
	}
	s := &Service{client: client, BasePath: basePath}
	s.Hostedmodels = NewHostedmodelsService(s)
	s.Training = NewTrainingService(s)
	return s, nil
	}

	type Service struct {
	client *http.Client
	BasePath string // API endpoint base URL

	Hostedmodels *HostedmodelsService

	Training *TrainingService
	}

	func NewHostedmodelsService(s Service) HostedmodelsService {
	rs := &HostedmodelsService{s: s}
	return rs
	}

	type HostedmodelsService struct {
	s *Service
	}

	func NewTrainingService(s Service) TrainingService {
	rs := &TrainingService{s: s}
	return rs
	}

	type TrainingService struct {
	s *Service
	}

	type Input struct {
	// Input: Input to the model for a prediction
	Input *InputInput `json:"input,omitempty"`
	}

	type InputInput struct {
	// CsvInstance: A list of input features, these can be strings or
	// doubles.
	CsvInstance []interface{} `json:"csvInstance,omitempty"`
	}

	type Output struct {
	// Id: The unique name for the predictive model.
	Id string `json:"id,omitempty"`

	// Kind: What kind of resource this is.
	Kind string `json:"kind,omitempty"`

	// OutputLabel: The most likely class [Categorical models only].
	OutputLabel string `json:"outputLabel,omitempty"`

	// OutputMulti: A list of classes with their estimated probabilities
	// [Categorical models only].
	OutputMulti []*OutputOutputMulti `json:"outputMulti,omitempty"`

	// OutputValue: The estimated regression value [Regression models only].
	OutputValue float64 `json:"outputValue,omitempty"`

	// SelfLink: A URL to re-request this resource.
	SelfLink string `json:"selfLink,omitempty"`
	}

	type OutputOutputMulti struct {
	// Label: The class label.
	Label string `json:"label,omitempty"`

	// Score: The probability of the class.
	Score float64 `json:"score,omitempty"`
	}

	type Training struct {
	// Id: The unique name for the predictive model.
	Id string `json:"id,omitempty"`

	// Kind: What kind of resource this is.
	Kind string `json:"kind,omitempty"`

	// ModelInfo: Model metadata.
	ModelInfo *TrainingModelInfo `json:"modelInfo,omitempty"`

	// SelfLink: A URL to re-request this resource.
	SelfLink string `json:"selfLink,omitempty"`

	// TrainingStatus: The current status of the training job. This can be
	// one of following: RUNNING; DONE; ERROR; ERROR: TRAINING JOB NOT FOUND
	TrainingStatus string `json:"trainingStatus,omitempty"`

	// Utility: A class weighting function, which allows the importance
	// weights for classes to be specified [Categorical models only].
	Utility []*TrainingUtility `json:"utility,omitempty"`
	}

	type TrainingModelInfo struct {
	// ClassWeightedAccuracy: Estimated accuracy of model taking utility
	// weights into account [Categorical models only].
	ClassWeightedAccuracy float64 `json:"classWeightedAccuracy,omitempty"`

	// ClassificationAccuracy: A number between 0.0 and 1.0, where 1.0 is
	// 100% accurate. This is an estimate, based on the amount and quality
	// of the training data, of the estimated prediction accuracy. You can
	// use this is a guide to decide whether the results are accurate enough
	// for your needs. This estimate will be more reliable if your real
	// input data is similar to your training data [Categorical models
	// only].
	ClassificationAccuracy float64 `json:"classificationAccuracy,omitempty"`

	// ConfusionMatrix: An output confusion matrix. This shows an estimate
	// for how this model will do in predictions. This is first indexed by
	// the true class label. For each true class label, this provides a pair
	// {predicted_label, count}, where count is the estimated number of
	// times the model will predict the predicted label given the true
	// label. Will not output if more then 100 classes [Categorical models
	// only].
	ConfusionMatrix *TrainingModelInfoConfusionMatrix `json:"confusionMatrix,omitempty"`

	// ConfusionMatrixRowTotals: A list of the confusion matrix row totals
	ConfusionMatrixRowTotals *TrainingModelInfoConfusionMatrixRowTotals `json:"confusionMatrixRowTotals,omitempty"`

	// MeanSquaredError: An estimated mean squared error. The can be used to
	// measure the quality of the predicted model [Regression models only].
	MeanSquaredError float64 `json:"meanSquaredError,omitempty"`

	// ModelType: Type of predictive model (CLASSIFICATION or REGRESSION)
	ModelType string `json:"modelType,omitempty"`

	// NumberClasses: Number of classes in the trained model [Categorical
	// models only].
	NumberClasses int64 `json:"numberClasses,omitempty,string"`

	// NumberInstances: Number of valid data instances used in the trained
	// model.
	NumberInstances int64 `json:"numberInstances,omitempty,string"`
	}

	type TrainingModelInfoConfusionMatrix struct {
	}

	type TrainingModelInfoConfusionMatrixRowTotals struct {
	}

	type TrainingUtility struct {
	}

	type Update struct {
	// ClassLabel: The true class label of this instance
	ClassLabel string `json:"classLabel,omitempty"`

	// CsvInstance: The input features for this instance
	CsvInstance []interface{} `json:"csvInstance,omitempty"`
	}

	// method id "prediction.hostedmodels.predict":

	type HostedmodelsPredictCall struct {
	s *Service
	hostedModelName string
	input *Input
	opt_ map[string]interface{}
	}

	// Predict: Submit input and request an output against a hosted model
	func (r HostedmodelsService) Predict(hostedModelName string, input Input) *HostedmodelsPredictCall {
	c := &HostedmodelsPredictCall{s: r.s, opt_: make(map[string]interface{})}
	c.hostedModelName = hostedModelName
	c.input = input
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c HostedmodelsPredictCall) Fields(s ...googleapi.Field) HostedmodelsPredictCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c HostedmodelsPredictCall) Do() (Output, error) {
	var body io.Reader = nil
	body, err := googleapi.WithoutDataWrapper.JSONReader(c.input)
	if err != nil {
	return nil, err
	}
	ctype := "application/json"
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "hostedmodels/{hostedModelName}/predict")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("POST", urls, body)
	googleapi.Expand(req.URL, map[string]string{
	"hostedModelName": c.hostedModelName,
	})
	req.Header.Set("Content-Type", ctype)
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return nil, err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return nil, err
	}
	var ret *Output
	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
	return nil, err
	}
	return ret, nil
	// {
	// "description": "Submit input and request an output against a hosted model",
	// "httpMethod": "POST",
	// "id": "prediction.hostedmodels.predict",
	// "parameterOrder": [
	// "hostedModelName"
	// ],
	// "parameters": {
	// "hostedModelName": {
	// "description": "The name of a hosted model",
	// "location": "path",
	// "required": true,
	// "type": "string"
	// }
	// },
	// "path": "hostedmodels/{hostedModelName}/predict",
	// "request": {
	// "$ref": "Input"
	// },
	// "response": {
	// "$ref": "Output"
	// },
	// "scopes": [
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}

	// method id "prediction.training.delete":

	type TrainingDeleteCall struct {
	s *Service
	data string
	opt_ map[string]interface{}
	}

	// Delete: Delete a trained model
	func (r TrainingService) Delete(data string) TrainingDeleteCall {
	c := &TrainingDeleteCall{s: r.s, opt_: make(map[string]interface{})}
	c.data = data
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c TrainingDeleteCall) Fields(s ...googleapi.Field) TrainingDeleteCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c *TrainingDeleteCall) Do() error {
	var body io.Reader = nil
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("DELETE", urls, body)
	googleapi.Expand(req.URL, map[string]string{
	"data": c.data,
	})
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return err
	}
	return nil
	// {
	// "description": "Delete a trained model",
	// "httpMethod": "DELETE",
	// "id": "prediction.training.delete",
	// "parameterOrder": [
	// "data"
	// ],
	// "parameters": {
	// "data": {
	// "description": "mybucket/mydata resource in Google Storage",
	// "location": "path",
	// "required": true,
	// "type": "string"
	// }
	// },
	// "path": "training/{data}",
	// "scopes": [
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}

	// method id "prediction.training.get":

	type TrainingGetCall struct {
	s *Service
	data string
	opt_ map[string]interface{}
	}

	// Get: Check training status of your model
	func (r TrainingService) Get(data string) TrainingGetCall {
	c := &TrainingGetCall{s: r.s, opt_: make(map[string]interface{})}
	c.data = data
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c TrainingGetCall) Fields(s ...googleapi.Field) TrainingGetCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c TrainingGetCall) Do() (Training, error) {
	var body io.Reader = nil
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("GET", urls, body)
	googleapi.Expand(req.URL, map[string]string{
	"data": c.data,
	})
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return nil, err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return nil, err
	}
	var ret *Training
	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
	return nil, err
	}
	return ret, nil
	// {
	// "description": "Check training status of your model",
	// "httpMethod": "GET",
	// "id": "prediction.training.get",
	// "parameterOrder": [
	// "data"
	// ],
	// "parameters": {
	// "data": {
	// "description": "mybucket/mydata resource in Google Storage",
	// "location": "path",
	// "required": true,
	// "type": "string"
	// }
	// },
	// "path": "training/{data}",
	// "response": {
	// "$ref": "Training"
	// },
	// "scopes": [
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}

	// method id "prediction.training.insert":

	type TrainingInsertCall struct {
	s *Service
	training *Training
	opt_ map[string]interface{}
	}

	// Insert: Begin training your model
	func (r TrainingService) Insert(training Training) *TrainingInsertCall {
	c := &TrainingInsertCall{s: r.s, opt_: make(map[string]interface{})}
	c.training = training
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c TrainingInsertCall) Fields(s ...googleapi.Field) TrainingInsertCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c TrainingInsertCall) Do() (Training, error) {
	var body io.Reader = nil
	body, err := googleapi.WithoutDataWrapper.JSONReader(c.training)
	if err != nil {
	return nil, err
	}
	ctype := "application/json"
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "training")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("POST", urls, body)
	googleapi.SetOpaque(req.URL)
	req.Header.Set("Content-Type", ctype)
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return nil, err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return nil, err
	}
	var ret *Training
	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
	return nil, err
	}
	return ret, nil
	// {
	// "description": "Begin training your model",
	// "httpMethod": "POST",
	// "id": "prediction.training.insert",
	// "path": "training",
	// "request": {
	// "$ref": "Training"
	// },
	// "response": {
	// "$ref": "Training"
	// },
	// "scopes": [
	// "https://www.googleapis.com/auth/devstorage.full_control",
	// "https://www.googleapis.com/auth/devstorage.read_only",
	// "https://www.googleapis.com/auth/devstorage.read_write",
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}

	// method id "prediction.training.predict":

	type TrainingPredictCall struct {
	s *Service
	data string
	input *Input
	opt_ map[string]interface{}
	}

	// Predict: Submit data and request a prediction
	func (r TrainingService) Predict(data string, input Input) *TrainingPredictCall {
	c := &TrainingPredictCall{s: r.s, opt_: make(map[string]interface{})}
	c.data = data
	c.input = input
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c TrainingPredictCall) Fields(s ...googleapi.Field) TrainingPredictCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c TrainingPredictCall) Do() (Output, error) {
	var body io.Reader = nil
	body, err := googleapi.WithoutDataWrapper.JSONReader(c.input)
	if err != nil {
	return nil, err
	}
	ctype := "application/json"
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}/predict")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("POST", urls, body)
	googleapi.Expand(req.URL, map[string]string{
	"data": c.data,
	})
	req.Header.Set("Content-Type", ctype)
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return nil, err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return nil, err
	}
	var ret *Output
	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
	return nil, err
	}
	return ret, nil
	// {
	// "description": "Submit data and request a prediction",
	// "httpMethod": "POST",
	// "id": "prediction.training.predict",
	// "parameterOrder": [
	// "data"
	// ],
	// "parameters": {
	// "data": {
	// "description": "mybucket/mydata resource in Google Storage",
	// "location": "path",
	// "required": true,
	// "type": "string"
	// }
	// },
	// "path": "training/{data}/predict",
	// "request": {
	// "$ref": "Input"
	// },
	// "response": {
	// "$ref": "Output"
	// },
	// "scopes": [
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}

	// method id "prediction.training.update":

	type TrainingUpdateCall struct {
	s *Service
	data string
	update *Update
	opt_ map[string]interface{}
	}

	// Update: Add new data to a trained model
	func (r TrainingService) Update(data string, update Update) *TrainingUpdateCall {
	c := &TrainingUpdateCall{s: r.s, opt_: make(map[string]interface{})}
	c.data = data
	c.update = update
	return c
	}

	// Fields allows partial responses to be retrieved.
	// See https://developers.google.com/gdata/docs/2.0/basics#PartialResponse
	// for more information.
	func (c TrainingUpdateCall) Fields(s ...googleapi.Field) TrainingUpdateCall {
	c.opt_["fields"] = googleapi.CombineFields(s)
	return c
	}

	func (c TrainingUpdateCall) Do() (Training, error) {
	var body io.Reader = nil
	body, err := googleapi.WithoutDataWrapper.JSONReader(c.update)
	if err != nil {
	return nil, err
	}
	ctype := "application/json"
	params := make(url.Values)
	params.Set("alt", "json")
	if v, ok := c.opt_["fields"]; ok {
	params.Set("fields", fmt.Sprintf("%v", v))
	}
	urls := googleapi.ResolveRelative(c.s.BasePath, "training/{data}")
	urls += "?" + params.Encode()
	req, _ := http.NewRequest("PUT", urls, body)
	googleapi.Expand(req.URL, map[string]string{
	"data": c.data,
	})
	req.Header.Set("Content-Type", ctype)
	req.Header.Set("User-Agent", "google-api-go-client/0.5")
	res, err := c.s.client.Do(req)
	if err != nil {
	return nil, err
	}
	defer googleapi.CloseBody(res)
	if err := googleapi.CheckResponse(res); err != nil {
	return nil, err
	}
	var ret *Training
	if err := json.NewDecoder(res.Body).Decode(&ret); err != nil {
	return nil, err
	}
	return ret, nil
	// {
	// "description": "Add new data to a trained model",
	// "httpMethod": "PUT",
	// "id": "prediction.training.update",
	// "parameterOrder": [
	// "data"
	// ],
	// "parameters": {
	// "data": {
	// "description": "mybucket/mydata resource in Google Storage",
	// "location": "path",
	// "required": true,
	// "type": "string"
	// }
	// },
	// "path": "training/{data}",
	// "request": {
	// "$ref": "Update"
	// },
	// "response": {
	// "$ref": "Training"
	// },
	// "scopes": [
	// "https://www.googleapis.com/auth/prediction"
	// ]
	// }

	}