topic Re: how to use JMP to predit the time? in Discussions

how to use JMP to predict the time?

lixia — Thu, 11 Dec 2025 15:05:19 GMT

i have three samples data like below, i want to predit the time of each sample through weibull model when the retention is below 70%.

time sample retention

1 1 100%

2 1 99.8%

....

298 1 81.9%

299 1 81%

300 1 80%

Re: how to use JMP to predit the time?

jthi — Thu, 11 Dec 2025 08:14:18 GMT

What have you tried? Reliability and Survival Methods (https://www.jmp.com/support/help/), Basic Analysis > Distributions > Options for Continuous Variables > Fit Distributions

Re: how to use JMP to predit the time?

christian-z — Thu, 11 Dec 2025 09:37:45 GMT

I recommend starting with the Degradation platform: Degradation

Edit: Specifically, this is called Inverse Prediction.

Re: how to use JMP to predit the time?

lixia — Fri, 12 Dec 2025 01:38:05 GMT

thank you for your reply. there is no weibull prediction in Degradation, and i want to set range to predict, for example, 250-300 hours data to predict the following hours.

Re: how to use JMP to predit the time?

lixia — Fri, 12 Dec 2025 01:41:23 GMT

thank you for you reply. i didn't find basic analysis in Reliability and Survival Methods. would you like to send me a screenshot？

-Jarmo

Re: how to use JMP to predit the time?

christian-z — Wed, 17 Dec 2025 12:51:01 GMT

Thanks for the clarification.

To set a range, I recommend using a Local Data Filter. This will allow you to just use the data from 250-300 hours. Almost all JMP platforms, including Degradation, update in real time as you change the filter. There are a few places you can click to find the Local Data Filter, but it is always in the Red Triangle Menu if you cannot find it.

Regarding the Weibull model - could you clarify a bit more, what exactly you want to use it for? Your data has the form "Retention vs Time". The Degradation platform will allow you to fit the trend of Retention vs Time, so you can extrapolate a Time to Failure for each sample. (If you cannot get a good fit with Degradation, maybe because your degradation path is nonlinear, try Repeated Measures Degradation).

Once you have your TTF for each sample, the next step in your analysis could be to do a Life Distribution, where you will find the Weibull distribution as well as some other commonly used distributions.

The Weibull model is also relevant for Survival Analysis which I do not think directly applies to the data format you provided, but might also be interesting for context (see: Exponential, Weibull, and Lognormal Plots and Fits).

JMP has a number of different platforms for analyzing Lifetime, Time to Failure, Degradation, and other related but different topics. This article has some explanations of the different terms JMP uses for these types of data: Reliability and Survival Methods

Re: how to use JMP to predit the time?

lixia — Thu, 18 Dec 2025 02:30:02 GMT

Thank you for your reply. I will try and reply.