IBM Watson Data & AI - Structured Ideas

Welcome to the idea forum for IBM Watson Data & AI — our team welcomes any feedback, requests, and suggestions you have for improving our products! 

This forum allows us to connect your product improvement ideas with IBM product and engineering teams. 

Ease call, edit, debug, and monitor OPL / CPLEX models within DSX

OPL is a much simpler way than Python for optimization modeling and today we provide only

https://twitter.com/AlexFleischer1/status/909759411484274689

 

If we had a simpler way OPL / CPLEX customers could move their models to DSX and DSX data scientists could move to optimization faster. This would help DSX compete against companies like Dataiku

 

NB: Some figures at https://www.linkedin.com/pulse/which-aml-gpl-do-you-use-call-cplex-alex-fleischer/

 

  • Alex Fleischer
  • Dec 20 2017
  • Needs review
Role Summary
  • Attach files
  • stephen Hall commented
    December 20, 2017 10:27

    Agreed, building an optimization model in Python required medium knowledge in optimization and programming. The barriere to entry is much lower with OPL. It would be great to be able to call and use an OPL model in a notebook.

  • Ferenc Katai commented
    December 20, 2017 12:11

    in addition, there are large assets in term of model development out there - especially in production -, so any new platform should target those customer-base AND if the client finds the python modeling more rewarding than anything else then they can still decide to move their assets to python (or any other language) and stay with the platform. Otherwise, either the platform excludes those clients or force them to move, which may not be a good start considering customer experience...

  • Martin Shell commented
    February 23, 2018 18:20

    I think this is a great idea. Perhaps we could have something similar to the way R Studio is embedded in DSX. Also would be great to leverage the data connectivity, facilitating the import of csv/RDBMS connections

  • Cesar G commented
    March 9, 2018 14:55

    You are correct Alex. The integration is the correct path for the tuple <opl,cplex,python>. I cannot still believe a design of DSX without OPL...indeed, I would recommend to IBM purchase AMPL and GAMS and make more robust OPL as a unique modeling tool

  • John Cui commented
    May 30, 2018 08:02

    OPL has a lot of fans, if DSX support OPL programming on local, it will connect with them to DSX, it is very important for both DSX and OPL.