MatthewReid854/reliability

Is Numerical Integration method available for stress-strength interference?

Closed this issue · 2 comments

In the tutorial there was only monte-carlo method to simulate the result:
tutorial link
However monte-carlo method might be slow , and the result may not be accurate. So I am wondering if numerical intergration method is available. Thanks for your help!

It seems not to difficult to replace the Monte-Carlo integration with another method.
By the way, the Monte-Carlo integration is also an numerical integration method like all others ;)

If all distributions under "reliability/Distributions.py" become an method to get the pdf and cdf like the scipy.stats api, there are a simple way to implement the feature.
I had it figured out and start an implementation under #5

Thanks for the suggestion Thomas (TEFEdotCC). I won't be pulling this into the main code but I have used a tiny part of your code (the bit that does the integration) and put into the main code. This has been used to replace the Monte Carlo method entirely as there is no point in keeping a slightly less accurate method.
I am currently working on one of the largest updates to reliability that I have done for a while. This will be part of the next release (V0.5.0) which should be ready in 1-2 weeks. I want to get these confidence intervals right before I release it.
I am now closing this issue as you can consider it implemented. I will credit you in readthedocs for the formula.