Hi, jessegrabowski, thank you very much for your help! I’ve looked into your suggestion, and I also found a similar issue in this post regarding custom Theano Ops for numerical integration: Custom Theano Op to do Numerical Integration – but it seems not entirely identical to my problem. My integration domain corresponds to the definition domain of a Gaussian process, where ‘r’ is the numerical domain over which I need to evaluate ‘f’.
In contrast, the integration domain in that post appears unrelated to their random variable. I’m a bit uncertain about how to approach such a situation.
Additionaly I’m not very familiar with PyTensor, which makes it challenging for me. However, I’m still willing to invest time and follow the advice you’ve provided to try and understand it.
If you have any further suggestions or guidance, I would greatly appreciate it. Once again, thank you for your help!