Reliability and Survival distribution graph by group

FlyDataFly

In the reliabilty and survival analysis, when running a life distribution fit, we can select to do the fitting by groups, and the life distribution report will show a graph with the different groups of data in different colors. Is there a way to sub-select which groups to display? The graph gets busy very quickly with more than a couple groups, but the displayed information is valuable for groups of interest in an analysis, so it would be useful to have a way to "turn off" the data for uninteresting groups or depending on the statistics observed in the summary table provided under the graph.

gzmorgan0

I am not an expert with Life Distribution, but I see three possible solutions:

1. Exclude and hide uninteresting groups
2. Aggregate uninteresting groups or create new groups
3. Add a Data Filter (Main Menu >Rows >Data Filter and select your Grouping column. Now you can select which groups to view and even automate the viewing.

With your data table current, create a table summary by your By Group. Select the uninteresting groups and exclude and hide and rerun the analysis. Another option is to create a new Grouping column with the groups of interest and the remaining groups are called Other, an aggregate distribution.

 

Suppose you were looking at life distributions of different tools, via the GUI select your grouping column, then from the main menu > Cols > select Recode to create your groups.  Your new groups can be used to exclude or create new subsetted tables.

 

A Data Filter can make viewing the individual goups a little easier. The jsl file adds a data filter and a check box to a life Distribution. Uncheck the checkbox to remove the confidence bands and check  to restore them. I like to use Inverse, so when I select a Group it is not removed but the only one showing. The screen shot is showing only Group 2

Press the Clear button to see all.  The sript was tested in JMP 12, 13, 14.

 

LifeDistribution_DataFilter.jsl