How do I affect forecasted failure counts in JMP Reliability Forecast with Fix Parameter?
Jan 31, 2017 12:51 PM(1284 views)
My question concerns the JMP Reliability Forecast Report. I can enter production and failure data under the Dates Format tab and get good results from the Life Distribution. I find that the Weibull distribution has the best fit to the data and I get a good prediction for the failure counts.
I want to investigate the effect of changing the distribution parameters- the Weibull beta and alpha on the predicted returns, but fixing the parameters under the parameter estimate does not affect the returns. Can anyone suggest a way to do this?
I don't think this is possible within the platform, and would involve a lot of work without it. As you observed I'm sure, once you 'Fix Parameters' and press 'Update', the scope of the update is only the 'Fix Parameters' outline node: The forecasted failure counts are still linked to the best fitting model you have used. From one point of view this is reasonable behaviour, since the model used to forecast is at least 'best' in a well-defined sense.
If I'm correct, this would be a new feature request - You could try email@example.com and explain what you would like to do.