Integration of model execution engine containers with a model development environment
An analytic model generated in a data scientist model acquisition toolbench is received. An analytic model is translated using a structuring tool. An execution simulation for testing the translated analytic model is performed for executing the model in the data scientist model acquisition toolbench.
1. A system, comprising:
a processor configured to:
receive an analytic model generated in a data scientist model acquisition toolbench, wherein the analytic model incorporates design rules comprising an input schema for a model input, a language specific execution code point, and an output schema for a model output;
wherein the language specific execution code point comprises an action directive to start scoring;
translate an analytic model using a structuring tool at least in part by abstracting the analytic model from its operational execution environment at least in part by using the input schema, the action directive, and the output schema; and
perform an execution simulation on an infrastructure system for testing the translated analytic model for executing the model in the data scientist model acquisition toolbench, wherein the infrastructure system is at least one of the following: an on-premises hardware system, an in-office computer, a datacenter, an off-premises hardware system, and a cloud infrastructure platform; and
a memory coupled to the processor and configured to provide the processor with instructions.
2. The system of claim 1 , wherein the processor is further configured to export the translated analytic model for deployment in a production execution environment.
3. The system of claim 1 , wherein the processor is further configured to export the translated analytic model to a virtualized execution environment for an analytic engine.
4. The system of claim 1 , wherein the structuring tool is integrated into the data scientist model acquisition toolbench.
5. The system of claim 1 , wherein the execution simulator is integrated into the data scientist model acquisition toolbench.
6. The system of claim 1 , wherein model deploy tools are integrated into the data scientist model acquisition toolbench.
7. The system of claim 1 , wherein the processor is further configured to publish to export the analytic model to a model manager.
8. The system of claim 1 , wherein the processor is further configured to perform model life cycle tracking.
9. The system of claim 8 , wherein performing model life cycle tracking comprises using a graph database for tracking any changes to the model from development through deployment life cycle.
10. The system of claim 1 , wherein the processor is further configured to support a plug-in for a second data scientist model acquisition toolbench.
11. The system of claim 1 , wherein translating comprises an object-based abstraction of the analytic model.
12. The system of claim 1 , wherein the data scientist model acquisition toolbench comprises at least one of the following: Jupyter, Rstudio, Zeppelin, Cloudera, and Data Science Toolbench.
13. The system of claim 1 , wherein the data scientist model acquisition toolbench comprises tools for at least one of the following: execution simulation, automatic schema generation, structuring, and test.
14. The system of claim 1 , wherein the analytic model is written in at least one of the following: C, Python, Java, R, S, SAS, PFA, H2O, PMML, SPSS, and MATLAB.
15. The system of claim 1 , wherein translating includes using reasonable factoring.
16. The system of claim 1 , wherein translating includes using smart comments.
17. The system of claim 1 , wherein translating includes using smart comments for input schema and output schema.
18. The system of claim 1 , wherein translating includes using smart comments for initializing code and ignoring code.
19. A method, comprising:
receiving an analytic model generated in a data scientist model acquisition toolbench, wherein the analytic model incorporates design rules comprising an input schema for a model input, a language specific execution code point, and an output schema for a model output;
wherein the language specific execution code point comprises an action directive to start scoring;
translating an analytic model using a structuring tool at least in part by abstracting the analytic model from its operational execution environment at least in part by using the input schema, the action directive, and the output schema; and
performing an execution simulation on an infrastructure system for testing the translated analytic model for executing the model in the data scientist model acquisition toolbench, wherein the infrastructure system is at least one of the following: an on-premises hardware system, an in-office computer, a datacenter, an off-premises hardware system, and a cloud infrastructure platform.
20. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:
receiving an analytic model generated in a data scientist model acquisition toolbench, wherein the analytic model incorporates design rules comprising an input schema for a model input, a language specific execution code point, and an output schema for a model output;
wherein the language specific execution code point comprises an action directive to start scoring;
translating an analytic model using a structuring tool at least in part by abstracting the analytic model from its operational execution environment at least in part by using the input schema, the action directive, and output schema; and
performing an execution simulation on an infrastructure system for testing the translated analytic model for executing the model in the data scientist model acquisition toolbench, wherein the infrastructure system is at least one of the following: an on-premises hardware system, an in-office computer, a datacenter, an off-premises hardware system, and a cloud infrastructure platform.