Recursive Cokriging Surrogate Model Development

I need assistance in developing a recursive cokriging multi-fidelity surrogate model primarily for optimization purposes. The model will utilize simulated data as its primary data source. I am currently using Python to generate this simulated data. The method we would like to implement is based on Loic le gratiet and Garnier 2014 paper, see links below. Old R code is available from the original authors. We would like to use the method to approximate a high-fidelity reservoir simulator with a low-fidelity reservoir simulator.

https://josselin-garnier.org/wp-content/uploads/2014/09/recursivecokriging.pdf
https://www.dl.begellhouse.com/journals/52034eb04b657aea,2f7b99cc281f2702,4c83626c5e64207a.html


Key Requirements:
- Develop a recursive cokriging model for optimization.
- Utilize simulated data as the data source.
- Ensure compatibility with Python-generated data.

Ideal Skills and Experience:
- Proficiency in Gaussian processes and developing surrogate models.
- Experience with optimization techniques.
- Strong background in handling simulated data, coding and statistics
- Expertise in Python for data generation and model integration.
- Expertise in R.