nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Two accelerating factors for censored data?

Peter_Hersh

I have a customer with historical reliability data where they have two stress factors temperature and voltage. The have time until failure and if it has failed or is censored. They want to be able to get mean life at certain temperature and voltage. I used fit life by x and used temperature as the x and voltage as the by variable. They wanted to be able to look at voltages they had not run in the past. Any ideas on the best way to do this?

Accepted answers

jiancao

Fit Parametric Survival allows you to add multiple factors. But you would have to hand-code new columns in order to accommodate transformation on the factors such as Arrhenius.

All comments

jiancao

Fit Parametric Survival allows you to add multiple factors. But you would have to hand-code new columns in order to accommodate transformation on the factors such as Arrhenius.

DonMcCormack

Pete,

I'm attaching a file I use to show two accelerating factors (just happen to be Temp and Voltage). It's from Meeker & Escobar, Section 19.3.3. See table 19.6 for the results. They should match what is generated from the save scripts. There's a slight catch. Don't use the Transform hotspot in the Fit Model dialog to transform the Temp column, but create an Arrhenius transformed column in the table. Not sure why Arrhenius doesn't work when used as a transformation choice. As Jian previously mentions, you'll use Fit Parametric.

Don

Tantalum Capacitor.jmp

Peter_Hersh

Thanks Jian and Don. That worked great makes a nice model takes the temperature and voltage interaction into account.