1.
|
2.
|
3.
|
4.
|
5.
|
6.
|
Select Individual Best as the Distribution.
|
7.
|
Make sure that AICc is the Comparison Criterion.
|
8.
|
Click OK.
|
10.
|
Select Omit for bearing seal, belt, container throw, cord short, and engine fan (the causes with the fewest failures).
|
In the initial Compare Distributions report, the probability and time axes are linear. But suppose that you want to see distribution estimates on a Fréchet scale.1
1.
|
2.
|
In the Compare Distributions report, select Fréchet in the Scale column.
|
3.
|
Select Interval Type > Pointwise from the red triangle menu.
|
4.
|
Select SEV in the Scale column.
|
The nonparametric estimates no longer approximate a straight line (Nonparametric Estimates with a SEV Probability Scale). You now know that the SEV distribution is not appropriate.
1.
|
2.
|
4.
|
5.
|
6.
|
7.
|
8.
|
Click OK.
|
9.
|
Select Weibull for Distribution and Scale.
|
•
|
You have no failures (all observations are right-censored) and the preference Weibayes Only for Zero Failure Data is checked. Then the Weibayes report appears. See Weibayes Example for Data with No Failures.
|
•
|
1.
|
2.
|
3.
|
4.
|
5.
|
6.
|
Select Likelihood as the Confidence Interval Method.
|
7.
|
Click OK.
|
9.
|
Click Update.
|
1.
|
2.
|
3.
|
4.
|
5.
|
6.
|
Select Likelihood as the Confidence Interval Method.
|
7.
|
Click OK.
|
8.
|
Select the Weibull distribution in the Compare Distributions plot.
|
9.
|
Select Fix Parameter from the Parametric Estimate - Weibull red triangle menu.
|
10.
|
12.
|
Click Update.
|
Using different scales is sometimes referred to as drawing the distribution on different types of probability paper.