Skip to content

Latest commit

 

History

History
54 lines (48 loc) · 3.21 KB

TreatmentResultRep.md

File metadata and controls

54 lines (48 loc) · 3.21 KB

LaunchDarklyApi::TreatmentResultRep

Properties

Name Type Description Notes
treatment_id String The ID of the treatment [optional]
treatment_name String The name of the treatment [optional]
mean Float The average value of the variation in this sample. It doesn’t capture the uncertainty in the measurement, so it should not be the only measurement you use to make decisions. [optional]
data_mean Float The mean of the data, with no priors effecting the result. [optional]
data_std_dev Float The standard deviation of the data, with no priors effecting the result. [optional]
credible_interval CredibleIntervalRep [optional]
p_best Float The likelihood that this variation has the biggest effect on the primary metric. The variation with the highest probability is likely the best of the variations you're testing [optional]
relative_differences Array<RelativeDifferenceRep> Estimates of the relative difference between this treatment's mean and the mean of each other treatment [optional]
units Integer The number of units exposed to this treatment that have event values, including those that are configured to default to 0 [optional]
traffic Integer The number of units exposed to this treatment. [optional]
event_values_sum Float The sum of the event values for the units exposed to this treatment. [optional]
distribution Distribution [optional]
correlation Float The outcome-covariate correlation [optional]
standard_deviation_ratio Float The ratio of the outcome SD to covariate SD [optional]
covariate_imbalance Float The imbalance between the covariate mean for the arm and the covariate mean for the experiment [optional]
variance_reduction Float The reduction in variance resulting from CUPED [optional]
model String The model used to calculate the results. Parameters specific to this model will be defined under the field under the same name [optional]
bayesian_normal BayesianNormalStatsRep [optional]
bayesian_beta BayesianBetaBinomialStatsRep [optional]

Example

require 'launchdarkly_api'

instance = LaunchDarklyApi::TreatmentResultRep.new(
  treatment_id: 92b8354e-360e-4d67-8f13-fa6a46ca8077,
  treatment_name: variation 25% off,
  mean: 0.5432525951557093,
  data_mean: 2.464845,
  data_std_dev: 1.0E-7,
  credible_interval: null,
  p_best: 0.6083,
  relative_differences: [{&quot;fromTreatmentId&quot;:&quot;92b8354e-360e-4d67-8f13-fa6a46ca8077&quot;,&quot;lower&quot;:-0.13708601934659803,&quot;upper&quot;:0.42655970355712425}],
  units: 76,
  traffic: 332,
  event_values_sum: 6234,
  distribution: null,
  correlation: null,
  standard_deviation_ratio: null,
  covariate_imbalance: null,
  variance_reduction: 0.1,
  model: bayesianNormal,
  bayesian_normal: null,
  bayesian_beta: null
)