The scoring code generation generates a model in C, Python, SAS, SQL, ... I creates functions encapsulating the model which you can then call from your own code.
https://community.jmp.com/t5/Discovery-Summit-2016/Scoring-Outside-the-Box/ta-p/22381
https://community.jmp.com/t5/Discussions/JMP-for-Machine-Learning-and-Score-Code-Deployment/td-p/415...
See examples on the community and the web, and JMP documentation.
JMP scoring code generated before 18 will not run in JMP 18 within the embedded Python environment due to a namespace clash, the scoring code utilizes the package namespace 'jmp'. In 18 this has been changed to 'jmp_score' as the namespace to avoid conflict with the new 'jmp' namespace containing access to data tables and columns. If you are running outside of JMP generated models should work regardless of being generated in < 18 or > 18. JMP models generated with 18 should work fine in prior JMP environments, but prior models will not run in 18, without being regenerated in 18.
