Observation Object

/experiments/EXPERIMENT_ID/observations/OBSERVATION_ID

The observed data from a single trial in your experiment. Observations have a value which reflects the value of the metric you are trying to optimize. Optionally, an observation can have a Standard Deviation (value_stddev) that reflects the noise in your measurement. If your trial has no noise, or is deterministic, you should omit the value_stddev parameter to ensure that you do not receive the same Suggestions multiple times.

Fields

KeyTypeValue
assignmentsAssignmentsThe values that the parameters held during this trial, usually set from a Suggestion.
createdintThe timestamp (in seconds since epoch) that this observation was reported.
experimentstringThe id of the Experiment for this observation.
failedbooleanA boolean indicating whether this observation failed for some reason. SigOpt takes this into consideration when optimizing. It is invalid for both failed to be true, and value or values to be present. The default value is false.
idstringA unique ID for this observation.
metadataMetadataOptional user-provided object. See Using Metadata for more information.
suggestionstringThe id of the Suggestion that provided the assignments for this observation.
valuefloatThe observed metric value from this trial. Only available for single metric experiments.
value_stddevfloatThe standard deviation in value that was observed in this trial. Only available for single metric experiments.
valuesarray<Metric Evaluation>An array of Metric Evaluation objects recording the values and other associated information for each metric in a given observation. This is available for both single metric and multimetric experiments. Values appear in alphabetical order by the metric name.

Examples

    {
      "assignments": {
        "degree": 2,
        "gamma": 3.6,
        "kernel": "rbf"
      },
      "created": 1558043983,
      "experiment": "1",
      "failed": false,
      "id": "1",
      "metadata": null,
      "object": "observation",
      "suggestion": "1",
      "value": 1,
      "value_stddev": null,
      "values": [
        {
          "name": "Accuracy",
          "object": "metric_evaluation",
          "value": 1,
          "value_stddev": null
        }
      ]
    }
    {
      "assignments": {
        "degree": 2,
        "gamma": 3.6,
        "kernel": "rbf"
      },
      "created": 1558043983,
      "experiment": "1",
      "failed": true,
      "id": "2",
      "metadata": null,
      "object": "observation",
      "suggestion": "2",
      "value": null,
      "value_stddev": null
    }