#lang scribble/manual
@(require planet/scribble (for-label racket))
@title{Prediction API v1.5}
@margin-note{This documentation has been automatically generated
                     using information supplied by the
                     Google API Discovery service.}
Lets you access a cloud hosted machine learning service that makes it easy to build smart apps
@hyperlink["https://developers.google.com/prediction/docs/developer-guide" "Google documentation."]
@table-of-contents{}
@defmodule[gapi/macro]
@racket[(require-gapi-doc prediction.v1.5.js)]
@section{API Parameters}
The following optional keyword arguments may be passed to
@italic{all} functions for this web service:
@defproc[(_
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]

) jsexpr?]{
@margin-note{This is not actually a function. This is just using
                     Scribble's defproc form to list the optional
                     keyword arguments that may be passed
                     to @italic{all} functions for this service.}
@racket[alt]: Data format for the response.

@racket[fields]: Selector specifying which fields to include in a partial response.

@racket[key]: API key. Your API key identifies your project and provides you with API access, quota, and reports. Required unless you provide an OAuth 2.0 token.

@racket[oauth_token]: OAuth 2.0 token for the current user.

@racket[prettyPrint]: Returns response with indentations and line breaks.

@racket[quotaUser]: Available to use for quota purposes for server-side applications. Can be any arbitrary string assigned to a user, but should not exceed 40 characters. Overrides userIp if both are provided.

@racket[userIp]: IP address of the site where the request originates. Use this if you want to enforce per-user limits.


}
@section{Resources}

@subsection{hostedmodels}
@defproc[(prediction-hostedmodels-predict
[#:hostedModelName hostedModelName string?]
[#:input input string? 'N/A]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Submit input and request an output against a hosted model.

@racket[hostedModelName]: The name of a hosted model.

@racket[input]: Input to the model for a prediction

}

@subsection{trainedmodels}
@defproc[(prediction-trainedmodels-list
[#:maxResults maxResults string? 'N/A]
[#:pageToken pageToken string? 'N/A]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
List available models.

@racket[maxResults]: Maximum number of results to return

@racket[pageToken]: Pagination token

}

@defproc[(prediction-trainedmodels-get
[#:id id string?]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Check training status of your model.

@racket[id]: The unique name for the predictive model.

}

@defproc[(prediction-trainedmodels-insert
[#:id id string? 'N/A]
[#:created created string? 'N/A]
[#:kind kind string? 'N/A]
[#:selfLink selfLink string? 'N/A]
[#:modelInfo modelInfo string? 'N/A]
[#:storageDataLocation storageDataLocation string? 'N/A]
[#:storagePMMLLocation storagePMMLLocation string? 'N/A]
[#:storagePMMLModelLocation storagePMMLModelLocation string? 'N/A]
[#:trainingComplete trainingComplete string? 'N/A]
[#:trainingStatus trainingStatus string? 'N/A]
[#:utility utility string? 'N/A]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Begin training your model.

@racket[id]: The unique name for the predictive model.

@racket[created]: Insert time of the model (as a RFC 3339 timestamp).

@racket[kind]: What kind of resource this is.

@racket[selfLink]: A URL to re-request this resource.

@racket[modelInfo]: Model metadata.

@racket[storageDataLocation]: Google storage location of the training data file.

@racket[storagePMMLLocation]: Google storage location of the preprocessing pmml file.

@racket[storagePMMLModelLocation]: Google storage location of the pmml model file.

@racket[trainingComplete]: Training completion time (as a RFC 3339 timestamp).

@racket[trainingStatus]: The current status of the training job. This can be one of following: RUNNING; DONE; ERROR; ERROR: TRAINING JOB NOT FOUND

@racket[utility]: A class weighting function, which allows the importance weights for class labels to be specified [Categorical models only].

}

@defproc[(prediction-trainedmodels-predict
[#:id id string?]
[#:input input string? 'N/A]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Submit model id and request a prediction.

@racket[id]: The unique name for the predictive model.

@racket[input]: Input to the model for a prediction

}

@defproc[(prediction-trainedmodels-analyze
[#:id id string?]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Get analysis of the model and the data the model was trained on.

@racket[id]: The unique name for the predictive model.

}

@defproc[(prediction-trainedmodels-delete
[#:id id string?]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Delete a trained model.

@racket[id]: The unique name for the predictive model.

}

@defproc[(prediction-trainedmodels-update
[#:id id string?]
[#:label label string? 'N/A]
[#:csvInstance csvInstance string? 'N/A]
[#:alt alt string? 'N/A]
[#:fields fields string? 'N/A]
[#:key key string? (api-key)]
[#:oauth_token oauth_token string? 'N/A]
[#:prettyPrint prettyPrint string? 'N/A]
[#:quotaUser quotaUser string? 'N/A]
[#:userIp userIp string? 'N/A]
) jsexpr?]{
Add new data to a trained model.

@racket[id]: The unique name for the predictive model.

@racket[label]: The class label of this instance

@racket[csvInstance]: The input features for this instance

}