WeibullAnalysisOptions Class

From ReliaWiki
Revision as of 18:05, 20 February 2014 by Chris Kahn (talk | contribs)
Jump to navigation Jump to search


Sets the control panel settings for use with the Calculate method.

Properties

Life Distribution

  • Distribution (type: WeibullSolverDistribution) Sets the life distribution to fit to the data set.
  • Parameters (type: WeibullSolverNumParameters) Sets the form of the selected life distribution (e.g., 2-parameter Weibull vs. mixed Weibull).
  • Subpopulations (type: integer) Sets the number of subpopulations in the mixed data set (mixed Weibull analysis only).

Control Panel Analysis Page Settings

  • Analysis (type: WeibullSolverMethod) Sets the method for estimating the parameters of the selected distribution (e.g., rank regression on X).
  • RankingMethod (type: WeibullSolverRankMethod) Sets the method for calculating the unreliability estimates of the times-to-failure data (e.g., median ranks).
  • UseRSRegression (type: boolean) Sets whether the ReliaSoft ranking method (RRM) will be used to calculate the unreliability estimates for times-to-failure data.
  • ConfBounds (type: boolean) Sets the method for calculating the confidence bounds (e.g., Fisher Matrix).
  • SortBeforeCalculations (type: boolean) Sets whether times-to-failure data will be sorted in ascending order before calculation.
  • UngroupGroupedData (type: boolean) Sets whether grouped data will be ungrouped when using rank regression.
  • Weibull_UnbiasParameters (type: boolean) Sets whether the biased beta estimate (due to MLE sampling error) will be corrected for the 2-parameter Weibull distribution.

Bayesian-Weibull Model Settings

  • Bayesian_PriorDistribution (type: WeibullBayesianPriorDistribution) Sets the distribution to be used as the prior distribution of beta for Bayesian-Weibull analysis.
  • Bayesian_Mean (type: double) Sets the mean parameter to use for Bayesian-Weibull analysis when the normal distribution is assumed.
  • Bayesian_Std (type: double) Sets the standard deviation parameter to use for Bayesian-Weibull analysis when the normal distribution is assumed.
  • Bayesian_Lambda (type: double) Sets the lambda parameter to use for Bayesian-Weibull analysis when the exponential distribution is assumed.
  • Bayesian_MinBeta (type: double) Sets the minimum beta parameter to use for Bayesian-Weibull analysis when a uniform distribution is assumed.
  • Bayesian_MaxBeta (type: double) Sets the maximum beta parameter to use for Bayesian-Weibull analysis when the assumed distribution is uniform.
  • Bayesian_Results (type: WeibullBayesianResultsOption) Sets whether the point estimates for beta are calculated from the median or mean value of the posterior distribution.

Application Setup Option

  • UseSpecialSort (type: boolean) Sets whether failures will always be put before suspensions when two identical times are encountered.

Usage Example

Declare a WeibullDataSet object.

       Dim WDS as New WeibullDataSet

Set Weibull as the distribution for analysis.

       WDS.AnalysisSettings.Distribution = WeibullSolverDistribution.Weibull

Specify that the 2-parameter version of the distribution will be used.

       WDS.AnalysisSettings.Parameters = WeibullSolverNumParameters.MS_2Parameter

Set MLE as the method that will be used to estimate the parameters.

       WDS.AnalysisSettings.Analysis = WeibullSolverMethod.MLE