cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Choose Language Hide Translation Bar
StatsHeadache
Level II

How to interpret parameters for weibull

I got this result in cumulative damage step stress - is b1 a threshold?

 

StatsHeadache_0-1711578824109.png

 

2 ACCEPTED SOLUTIONS

Accepted Solutions
StatsHeadache
Level II

Re: How to interpret parameters for weibull

Ok wait, maybe yes!  looks like the equation below matches and X is number of events? creating mu, which would be eta in the weibull distribution?

StatsHeadache_0-1711642526636.png

 

View solution in original post

7 REPLIES 7
txnelson
Super User

Re: How to interpret parameters for weibull

Try going to the Tools and clicking on the "?" and then click on the table and it will display the help page for the report table.

Jim

Re: How to interpret parameters for weibull

Are you using an Arrhenius transform? I think b0 and b1 are the intercept and slope for the linearization of the Alpha parameter of the Weibull distribution.

StatsHeadache
Level II

Re: How to interpret parameters for weibull

No its a linear prediction

StatsHeadache
Level II

Re: How to interpret parameters for weibull

Ok wait, maybe yes!  looks like the equation below matches and X is number of events? creating mu, which would be eta in the weibull distribution?

StatsHeadache_0-1711642526636.png

 

peng_liu
Staff

Re: How to interpret parameters for weibull

X is your acceleration factor, e.g. temperature, voltage, etc.

If your "eta" is "alpha" on this page Statistical Details for Distributions , then exp(mu) is your "eta".

MRB3855
Super User

Re: How to interpret parameters for weibull

StatsHeadache
Level II

Re: How to interpret parameters for weibull

OK thats the best answer, but the slope is usually Beta, and perhaps intercept can be interpreted as threshold - that would work, that leave Eta usually described as scale factor as negative, which isn't possible.

Which leaves my best guess as:

b0 = eta

b1 = threshold (lambda)

beta = beta (slope)