The Open Science Community (OSC-W) is organising a lunch session (11. January, 12.30-13.30, Impulse) on modelling, how to maintain models after (PhD/European/...) projects ended and how we can further develop existing models. Daniel Reyes Lastiri is hosting this seminar.
We want to start a dialogue and brainstorm on the use of models and the role of different WUR stakeholders. The aim is to find solutions together, and perhaps take the first steps in pioneering a repository of open-source tools for agriculture. Further information below and on our website.
Event website and sign-up: https://openscience-wageningen.com/events/lunchdaniel/
Daniel’s personal experience in a nutshell:
I did my PhD on modelling aquaponics, a production method used to grow fish and plants together. For a number of reasons, I found myself making models from scratch, because:
• models in literature were too complex and detailed in the physiology for my simple system-level interests,
• the description in paper was limited or contained mistakes that made it difficult to implement in code,
• the code was available, but its documentation was limited and hard to follow.
Honestly, my own PhD code is hard to follow, as I’m still learning how to properly document it.
I’m now teaching courses on mathematical modelling, and I’ve noticed that the best way to put my code to the test is to have dozens of students applying it every year. The code we use when teaching is becoming a very reliable tool.
Similar to the teaching experience, I wish we could have time or specialized staff dedicated to maintaining the code from our research projects, so that we could use it for education and further research. And I wish we had a central platform to actively engage in maintenance and development of the models at WUR. But it is hard to find funds to allocate working hours or full-time staff to activities that do not produce direct research outputs.
There is already a good model library at WUR from Plant Production Systems. But it’s just a library. I’m talking about a live repository where the community of users can contribute continuously to the development by reporting bugs and improvements. For example via Git, first within WUR and later world-wide.