Special Cases of Life Data Analysis (LDA)
While most of the information presented in this section is specific to parametric LDA, special cases of LDA are also supported. Any data set that must be calculated in some special way is considered a special case. This includes data sets for non-parametric and reliability growth analysis as well as data sets for warranty and degradation analysis.
Data sets for non-parametric and reliability growth analysis are special cases because they are analyzed independently of the calculations used for distributions. Data sets for warranty and degradation data sets are special cases because they require special data formats for capturing the data to analyze.
• You use non-parametric LDA when you do not want to assume a distribution or when you are calculating a probability at each observed failure time. For more information, see
Non-Parametric Life Data Analysis (LDA).
• You use reliability growth analysis to track the growth of the reliability of a product over time, especially during development. This iterative process uses either the Duane or Crow/AMSAA model to calculate parameter values. For more information, see
Reliability Growth Analysis.
• You use warranty analysis to forecast future failure quantities based on information that you supply about the date a product was produced or sold and the date it was returned from the field. For more information, see
Warranty Analysis.
• You use degradation analysis to predict future failure times based on information that you supply about the changing rate of degradation for a product and the critical value at which the unit is assumed to have failed. For more information, see
Degradation Analysis.
For all of these special cases but reliability growth, you can save the results calculated to a new or existing parametric data set in the free-form format so that you can then analyze it using distributions. For more information, see
Resulting Data Sets.