IMPORTANT: No additional bug fixes or documentation updates will be released for this version. For the latest information, see the current release documentation.

« Get data frame analytics jobs API Get trained models API »

› › ›

Get data frame analytics jobs statistics API

edit

Get data frame analytics jobs statistics API

edit

Retrieves usage information for data frame analytics jobs.

This functionality is in beta and is subject to change. The design and code is less mature than official GA features and is being provided as-is with no warranties. Beta features are not subject to the support SLA of official GA features.

Request

edit

GET _ml/data_frame/analytics/<data_frame_analytics_id>/_stats

GET _ml/data_frame/analytics/<data_frame_analytics_id>,<data_frame_analytics_id>/_stats

GET _ml/data_frame/analytics/_stats

GET _ml/data_frame/analytics/_all/_stats

GET _ml/data_frame/analytics/*/_stats

Prerequisites

edit

If the Elasticsearch security features are enabled, you must have the following privileges:

cluster: monitor_ml

For more information, see Security privileges and Machine learning security privileges.

Path parameters

edit

<data_frame_analytics_id>: (Optional, string) Identifier for the data frame analytics job. If you do not specify this option, the API returns information for the first hundred data frame analytics jobs.

Query parameters

edit

allow_no_match

(Optional, Boolean) Specifies what to do when the request:

Contains wildcard expressions and there are no data frame analytics jobs that match.
Contains the _all string or no identifiers and there are no matches.
Contains wildcard expressions and there are only partial matches.

The default value is true, which returns an empty data_frame_analytics array when there are no matches and the subset of results when there are partial matches. If this parameter is false, the request returns a 404 status code when there are no matches or only partial matches.

from

(Optional, integer) Skips the specified number of data frame analytics jobs. The default value is 0.

size

(Optional, integer) Specifies the maximum number of data frame analytics jobs to obtain. The default value is 100.

verbose

(Optional, Boolean) Defines whether the stats response should be verbose. The default value is false.

Response body

edit

data_frame_analytics

(array) An array of objects that contain usage information for data frame analytics jobs, which are sorted by the id value in ascending order.

Properties of data frame analytics job usage resources

analysis_stats

(object) An object containing information about the analysis job.

Properties of analysis_stats

classification_stats

(object) An object containing information about the classification analysis job.

Properties of classification_stats

hyperparameters

(object) An object containing the parameters of the classification analysis job.

Properties of hyperparameters

alpha: (double) Advanced configuration option. Machine learning uses loss guided tree growing, which means that the decision trees grow where the regularized loss decreases most quickly. This parameter affects loss calculations by acting as a multiplier of the tree depth. Higher alpha values result in shallower trees and faster training times. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to zero.
class_assignment_objective: (string) Defines the objective to optimize when assigning class labels: maximize_accuracy or maximize_minimum_recall. When maximizing accuracy, class labels are chosen to maximize the number of correct predictions. When maximizing minimum recall, labels are chosen to maximize the minimum recall for any class. Defaults to maximize_minimum_recall.
downsample_factor: (double) Advanced configuration option. Controls the fraction of data that is used to compute the derivatives of the loss function for tree training. A small value results in the use of a small fraction of the data. If this value is set to be less than 1, accuracy typically improves. However, too small a value may result in poor convergence for the ensemble and so require more trees. For more information about shrinkage, refer to this wiki article. By default, this value is calculated during hyperparameter optimization. It must be greater than zero and less than or equal to 1.
eta: (double) Advanced configuration option. The shrinkage applied to the weights. Smaller values result in larger forests which have a better generalization error. However, larger forests cause slower training. For more information about shrinkage, refer to this wiki article. By default, this value is calculated during hyperparameter optimization. It must be a value between 0.001 and 1.
eta_growth_rate_per_tree: (double) Advanced configuration option. Specifies the rate at which eta increases for each new tree that is added to the forest. For example, a rate of 1.05 increases eta by 5% for each extra tree. By default, this value is calculated during hyperparameter optimization. It must be between 0.5 and 2.
feature_bag_fraction: (double) Advanced configuration option. Defines the fraction of features that will be used when selecting a random bag for each candidate split. By default, this value is calculated during hyperparameter optimization.
gamma: (double) Advanced configuration option. Regularization parameter to prevent overfitting on the training data set. Multiplies a linear penalty associated with the size of individual trees in the forest. A high gamma value causes training to prefer small trees. A small gamma value results in larger individual trees and slower training. By default, this value is calculated during hyperparameter optimization. It must be a nonnegative value.
lambda: (double) Advanced configuration option. Regularization parameter to prevent overfitting on the training data set. Multiplies an L2 regularization term which applies to leaf weights of the individual trees in the forest. A high lambda value causes training to favor small leaf weights. This behavior makes the prediction function smoother at the expense of potentially not being able to capture relevant relationships between the features and the dependent variable. A small lambda value results in large individual trees and slower training. By default, this value is calculated during hyperparameter optimization. It must be a nonnegative value.
max_attempts_to_add_tree: (integer) If the algorithm fails to determine a non-trivial tree (more than a single leaf), this parameter determines how many of such consecutive failures are tolerated. Once the number of attempts exceeds the threshold, the forest training stops.
max_optimization_rounds_per_hyperparameter: (integer) Advanced configuration option. A multiplier responsible for determining the maximum number of hyperparameter optimization steps in the Bayesian optimization procedure. The maximum number of steps is determined based on the number of undefined hyperparameters times the maximum optimization rounds per hyperparameter. By default, this value is calculated during hyperparameter optimization.
max_trees: (integer) Advanced configuration option. Defines the maximum number of decision trees in the forest. The maximum value is 2000. By default, this value is calculated during hyperparameter optimization.
num_folds: (integer) The maximum number of folds for the cross-validation procedure.
num_splits_per_feature: (integer) Determines the maximum number of splits for every feature that can occur in a decision tree when the tree is trained.
soft_tree_depth_limit: (double) Advanced configuration option. Machine learning uses loss guided tree growing, which means that the decision trees grow where the regularized loss decreases most quickly. This soft limit combines with the soft_tree_depth_tolerance to penalize trees that exceed the specified depth; the regularized loss increases quickly beyond this depth. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to 0.
soft_tree_depth_tolerance: (double) Advanced configuration option. This option controls how quickly the regularized loss increases when the tree depth exceeds soft_tree_depth_limit. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to 0.01.

iteration

(integer) The number of iterations on the analysis.

timestamp

(date) The timestamp when the statistics were reported in milliseconds since the epoch.

timing_stats

(object) An object containing time statistics about the data frame analytics job.

Properties of timing_stats

elapsed_time: (integer) Runtime of the analysis in milliseconds.
iteration_time: (integer) Runtime of the latest iteration of the analysis in milliseconds.

validation_loss

(object) An object containing information about validation loss.

Properties of validation_loss

fold_values: (array of strings) Validation loss values for every added decision tree during the forest growing procedure.
loss_type: (string) The type of the loss metric. For example, binomial_logistic.

outlier_detection_stats

(object) An object containing information about the outlier detection job.

Properties of outlier_detection_stats

parameters

(object) The list of job parameters specified by the user or determined by algorithmic heuristics.

Properties of parameters

compute_feature_influence: (Boolean) Specifies whether the feature influence calculation is enabled. Defaults to true.
feature_influence_threshold: (double) The minimum outlier score that a document needs to have to calculate its feature influence score. Value range: 0-1 (0.1 by default).
method: (string) The method that outlier detection uses. Available methods are lof, ldof, distance_kth_nn, distance_knn, and ensemble. The default value is ensemble, which means that outlier detection uses an ensemble of different methods and normalises and combines their individual outlier scores to obtain the overall outlier score.
n_neighbors: (integer) Defines the value for how many nearest neighbors each method of outlier detection uses to calculate its outlier score. When the value is not set, different values are used for different ensemble members. This deafault behavior helps improve the diversity in the ensemble; only override it if you are confident that the value you choose is appropriate for the data set.
outlier_fraction: (double) The proportion of the data set that is assumed to be outlying prior to outlier detection. For example, 0.05 means it is assumed that 5% of values are real outliers and 95% are inliers.
standardization_enabled: (Boolean) If true, the following operation is performed on the columns before computing outlier scores: (x_i - mean(x_i)) / sd(x_i). Defaults to true. For more information about this concept, see Wikipedia.

timestamp

(date) The timestamp when the statistics were reported in milliseconds since the epoch.

timing_stats

(object) An object containing time statistics about the data frame analytics job.

Property of timing_stats

elapsed_time: (integer) Runtime of the analysis in milliseconds.

regression_stats

(object) An object containing information about the regression analysis.

Properties of regression_stats

hyperparameters

(object) An object containing the parameters of the regression analysis.

Properties of hyperparameters

alpha: (double) Advanced configuration option. Machine learning uses loss guided tree growing, which means that the decision trees grow where the regularized loss decreases most quickly. This parameter affects loss calculations by acting as a multiplier of the tree depth. Higher alpha values result in shallower trees and faster training times. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to zero.
downsample_factor: (double) Advanced configuration option. Controls the fraction of data that is used to compute the derivatives of the loss function for tree training. A small value results in the use of a small fraction of the data. If this value is set to be less than 1, accuracy typically improves. However, too small a value may result in poor convergence for the ensemble and so require more trees. For more information about shrinkage, refer to this wiki article. By default, this value is calculated during hyperparameter optimization. It must be greater than zero and less than or equal to 1.
eta: (double) Advanced configuration option. The shrinkage applied to the weights. Smaller values result in larger forests which have a better generalization error. However, larger forests cause slower training. For more information about shrinkage, refer to this wiki article. By default, this value is calculated during hyperparameter optimization. It must be a value between 0.001 and 1.
eta_growth_rate_per_tree: (double) Advanced configuration option. Specifies the rate at which eta increases for each new tree that is added to the forest. For example, a rate of 1.05 increases eta by 5% for each extra tree. By default, this value is calculated during hyperparameter optimization. It must be between 0.5 and 2.
feature_bag_fraction: (double) Advanced configuration option. Defines the fraction of features that will be used when selecting a random bag for each candidate split. By default, this value is calculated during hyperparameter optimization.
gamma: (double) Advanced configuration option. Regularization parameter to prevent overfitting on the training data set. Multiplies a linear penalty associated with the size of individual trees in the forest. A high gamma value causes training to prefer small trees. A small gamma value results in larger individual trees and slower training. By default, this value is calculated during hyperparameter optimization. It must be a nonnegative value.
lambda: (double) Advanced configuration option. Regularization parameter to prevent overfitting on the training data set. Multiplies an L2 regularization term which applies to leaf weights of the individual trees in the forest. A high lambda value causes training to favor small leaf weights. This behavior makes the prediction function smoother at the expense of potentially not being able to capture relevant relationships between the features and the dependent variable. A small lambda value results in large individual trees and slower training. By default, this value is calculated during hyperparameter optimization. It must be a nonnegative value.
max_attempts_to_add_tree: (integer) If the algorithm fails to determine a non-trivial tree (more than a single leaf), this parameter determines how many of such consecutive failures are tolerated. Once the number of attempts exceeds the threshold, the forest training stops.
max_optimization_rounds_per_hyperparameter: (integer) Advanced configuration option. A multiplier responsible for determining the maximum number of hyperparameter optimization steps in the Bayesian optimization procedure. The maximum number of steps is determined based on the number of undefined hyperparameters times the maximum optimization rounds per hyperparameter. By default, this value is calculated during hyperparameter optimization.
max_trees: (integer) Advanced configuration option. Defines the maximum number of decision trees in the forest. The maximum value is 2000. By default, this value is calculated during hyperparameter optimization.
num_folds: (integer) The maximum number of folds for the cross-validation procedure.
num_splits_per_feature: (integer) Determines the maximum number of splits for every feature that can occur in a decision tree when the tree is trained.
soft_tree_depth_limit: (double) Advanced configuration option. Machine learning uses loss guided tree growing, which means that the decision trees grow where the regularized loss decreases most quickly. This soft limit combines with the soft_tree_depth_tolerance to penalize trees that exceed the specified depth; the regularized loss increases quickly beyond this depth. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to 0.
soft_tree_depth_tolerance: (double) Advanced configuration option. This option controls how quickly the regularized loss increases when the tree depth exceeds soft_tree_depth_limit. By default, this value is calculated during hyperparameter optimization. It must be greater than or equal to 0.01.

iteration

(integer) The number of iterations on the analysis.

timestamp

(date) The timestamp when the statistics were reported in milliseconds since the epoch.

timing_stats

(object) An object containing time statistics about the data frame analytics job.

Propertis of timing_stats

elapsed_time: (integer) Runtime of the analysis in milliseconds.
iteration_time: (integer) Runtime of the latest iteration of the analysis in milliseconds.

validation_loss

(object) An object containing information about validation loss.

Properties of validation_loss

fold_values: (array of strings) Validation loss values for every added decision tree during the forest growing procedure.
loss_type: (string) The type of the loss metric. For example, binomial_logistic.

assignment_explanation

(string) For running jobs only, contains messages relating to the selection of a node to run the job.

data_counts

(object) An object that provides counts for the quantity of documents skipped, used in training, or available for testing.

Properties of data_counts

skipped_docs_count: (integer) The number of documents that are skipped during the analysis because they contained values that are not supported by the analysis. For example, outlier detection does not support missing fields so it skips documents with missing fields. Likewise, all types of analysis skip documents that contain arrays with more than one element.
test_docs_count: (integer) The number of documents that are not used for training the model and can be used for testing.
training_docs_count: (integer) The number of documents that are used for training the model.

id

(string) The unique identifier of the data frame analytics job.

memory_usage

(Optional, object) An object describing memory usage of the analytics. It is present only after the job is started and memory usage is reported.

Properties of memory_usage

memory_reestimate_bytes

(long) This value is present when the status is hard_limit and it is a new estimate of how much memory the job needs.

peak_usage_bytes

(long) The number of bytes used at the highest peak of memory usage.

status

(string) The memory usage status. May have one of the following values:

ok: usage stayed below the limit.
hard_limit: usage surpassed the configured memory limit.

timestamp

(date) The timestamp when memory usage was calculated.

node

(object) Contains properties for the node that runs the job. This information is available only for running jobs.

Properties of node

attributes: (object) Lists node attributes such as ml.machine_memory or ml.max_open_jobs settings.
ephemeral_id: (string) The ephemeral ID of the node.
id: (string) The unique identifier of the node.
name: (string) The node name.
transport_address: (string) The host and port where transport HTTP connections are accepted.

progress

(array) The progress report of the data frame analytics job by phase.

Properties of phase objects

phase

(string) Defines the phase of the data frame analytics job. Possible phases:

reindexing,
loading_data,
computing_outliers (for outlier detection only),
feature_selection (for regression and classification only),
coarse_parameter_search (for regression and classification only),
fine_tuning_parameters (for regression and classification only),
final_training (for regression and classification only),
writing_results,
inference (for regression and classification only).

To learn more about the different phases, refer to How a data frame analytics job works.

progress_percent

(integer) The progress that the data frame analytics job has made expressed in percentage.

state

(string) The status of the data frame analytics job, which can be one of the following values: failed, started, starting,stopping, stopped.

Response codes

edit

404 (Missing resources): If allow_no_match is false, this code indicates that there are no resources that match the request or only partial matches for the request.

Examples

edit

The following API retrieves usage information for the outlier detection data frame analytics job example:

GET _ml/data_frame/analytics/ecommerce/_stats

The API returns the following results:

{
  "count" : 1,
  "data_frame_analytics" : [
    {
      "id" : "ecommerce",
      "state" : "stopped",
      "progress" : [
        {
          "phase" : "reindexing",
          "progress_percent" : 100
        },
        {
          "phase" : "loading_data",
          "progress_percent" : 100
        },
        {
          "phase" : "analyzing",
          "progress_percent" : 100
        },
        {
          "phase" : "writing_results",
          "progress_percent" : 100
        }
      ],
      "data_counts" : {
        "training_docs_count" : 3321,
        "test_docs_count" : 0,
        "skipped_docs_count" : 0
      },
      "memory_usage" : {
        "timestamp" : 1586905058000,
        "peak_usage_bytes" : 279484
      },
      "analysis_stats" : {
        "outlier_detection_stats" : {
          "timestamp" : 1586905058000,
          "parameters" : {
            "n_neighbors" : 0,
            "method" : "ensemble",
            "compute_feature_influence" : true,
            "feature_influence_threshold" : 0.1,
            "outlier_fraction" : 0.05,
            "standardization_enabled" : true
          },
          "timing_stats" : {
            "elapsed_time" : 245
          }
        }
      }
    }
  ]
}

« Get data frame analytics jobs API Get trained models API »