The EncoderStore API provides methods to store and retrieve encoders.

This API requires the Predictive Intelligence plugin (com.glide.platform_ml) and is provided within the sn_ml namespace.

EncoderStore - add(Object mlEncoder)

Adds a new encoder object to the store and returns a unique name.

Note: Label values do not need to be unique. For example, if you run this method with the same label 10 times, this method adds 10 different uniquely-named objects to the store.
Table 1. Parameters
Name Type Description
mlEncoder Encoder Encoder() object to add to the store.
Table 2. Returns
Type Description
String System-generated solution name.

Example

The following example shows how to add an encoder to the store. Use Encoder - submitTrainingJob() to run the training job after adding it to the store.

// Create a dataset 
var myData = new sn_ml.DatasetDefinition({

  'tableName' : 'incident',
  'fieldNames' : ['assignment_group', 'short_description', 'description'],
  'encodedQuery' : 'activeANYTHING'

});

// Create a solution 
var myEncoder = new sn_ml.Encoder({

  'label': "my encoder definition",
  'datasets' : [myData],
  'predictedFieldName' : 'assignment_group',
  'inputFieldNames':['short_description']

});

// Add the encoder to the store to later be able to retrieve it.
var my_unique_name = sn_ml.EncoderStore.add(myEncoder);

EncoderStore - deleteObject(String name)

Removes a specified encoder object from the store.

Table 3. Parameters
Name Type Description
name String Name of the Encoder() object to be deleted.
Table 4. Returns
Type Description
None

Example

The following example shows how to delete an encoder from the store.

sn_ml.EncoderStore.deleteObject("ml_sn_global_global_encoder");

EncoderStore - get(String name)

Gets an encoder object from a store.

Table 5. Parameters
Name Type Description
name String Name of an encoder in a store.
Table 6. Returns
Type Description
Object Encoder object. Returns an error if the object does not exist.

Example

The following example shows how to get an encoder object from the store using the get() method and view its training status using the Encoder - getActiveVersion() and EncoderVersion - getStatus() methods.

// Get status
var myEncoder = sn_ml.EncoderStore.get('ml_incident_categorization');

gs.print(JSON.stringify(JSON.parse(myEncoder.getActiveVersion().getStatus(), null, 2)));

Output:

{
 "state":"solution_complete",
 "percentComplete":"100",
 "hasJobEnded":"true"
}

EncoderStore - getAllNames(Object options)

Gets the names of all encoder definition records in the store.

Table 7. Parameters
NameTypeDescription
options Object Options for restricting results within the specified properties.
{
  "label": "String",
  "domainName": "String",
  "scope": "String"
};
options.label String Optional. Label of your solution object.
options.domainName String Optional. Name of the domain for your solution object. Refer to Domain separation and Predictive Intelligence.
options.scope String Optional. Name of an application scope for your solution object.
Table 8. Returns
Type Description
Array List of strings representing encoder object names in the store.

Example

In the following example, the getAllNames() method returns a list of all names in the store.

gs.print(JSON.stringify(JSON.parse(sn_ml.EncoderStore.getAllNames()), null, 2));

Output:

[
  "ml_x_snc_global_global_classification_word_corpus",
  "ml_x_snc_global_global_predictability_estimate",
  "GloVe",
  "ml_x_snc_global_global_encoder",
  "ml_x_snc_global_global_predictability_estimate_1"
]

Example

In the following example, the getAllNames() method returns only names associated with values set in the options parameter.

var options = {
  'label' : 'my encoder definition',
  'domainName' : 'global',
  'scope' : 'global'
};
var solNames = sn_ml.EncoderStore.getAllNames(options);
gs.print(JSON.stringify(JSON.parse(solNames), null, 2));

Output:

[
  "ml_x_snc_global_global_my_encoder_definition"
]

EncoderStore - update(String name, Object mlEncoder)

Updates an encoder object in a store.

Table 9. Parameters
Name Type Description
name String Name of the encoder to update.
mlEncoder Encoder Encoder() object properties to update.
Table 10. Returns
Type Description
None

Example

The following example shows how to update an encoder object in the store.

var encoderUpdate = new sn_ml.Encoder({
  'label': 'my encoder definition',
  'dataset' : myData,
  'predictedFieldName' : 'assignment_group',
  'inputFieldNames': ['short_description']
});

sn_ml.EncoderStore.update('ml_sn_global_global_incident_service', encoderUpdate);