ANALYZE

Syntax:

ANALYZE 'data' BY MODEL model_name USING method_name

Examples:

jubaql> ANALYZE '{"name": "慶喜"}' BY MODEL cls USING classify
{
  "predictions":[{
    "label":"徳川",
    "score":0.05624999850988388
  },{
    "label":"北条",
    "score":0.0
  },{
    "label":"足利",
    "score":0.0
  }]
}

jubaql> ANALYZE '荻野貴司' BY MODEL reco USING complete_row_from_id
{
  "string_values":{
  },
  "num_values":{
    "長打率":0.34623855352401733,
    "試合数":102.7064208984375,
    "打数":325.32110595703125,
    "安打":84.64220428466797,
    "RC27":3.8101837635040283,
    "出塁率":0.3158532977104187,
    "OPS":0.6621102094650269,
    "盗塁":6.31192684173584,
    "打率":0.25398167967796326,
    "四球":27.706422805786133,
    "打席":367.9449462890625,
    "打点":31.302751541137695,
    "併殺打":6.2110090255737305,
    "犠打":9.2660551071167,
    "死球":3.4770641326904297,
    "三振":54.40367126464844,
    "本塁打":4.917431354522705,
    "XR27":3.7999088764190674
  }
}

Explanation

ANALYZE queries a previously defined and trained Jubatus model for results of the learning process.

• data is a string that will become the parameter of the given Jubatus method. When this parameter is a datum, data is expected to be a JSON string and will be converted to a datum using the columns specified in the WITH clause of the CREATE MODEL statement.
• model_name is the name of a model previously defined using CREATE MODEL. (For the returned value to make any sense, that model should have also been trained using the UPDATE MODEL statement.)
• method_name is the method to use for analyzing the model:
  • calc_score for an ANOMALY model,
  • classify for a CLASSIFIER model,
  • complete_row_from_id or complete_row_from_datum for a RECOMMENDER model.

Notes

• The results will be returned as JSON corresponding to each method’s return type.