Examples of PythonGateway Usage. Python Gateway for InterSystems Data Platforms. Execute Python code and more from InterSystems IRIS brings you the power of Python right into your InterSystems IRIS environment:
do $system.OBJ.ImportDir("C:\InterSystems\Repos\PythongatewaySamples\","*.cls","c",,1)) into Production (Ensemble-enabled) namespace. In case you want to Production-enable namespace call:
write ##class(%EnsembleMgr).EnableNamespace($Namespace, 1).
pip install recordlinkage pip install dedupe pip install pandas pip install numpy pip install matplotlib pip install sklearn pip install seaborn
To choose specific example filter by production category.
We start with working self-correcting model used for predictive maintenance. First, we see how production elements work together and how our transactional processes can benefit from AI/ML models. After that we’d improve the model and see how this change propagates through production. Finally, we’ll explore different applications of this architecture.
We want to do predictive maintenance on engines. Dataset: we store information about one specific engine. We get a large array of sensor data every second. Historic data is already in the dataset. Check Engine.md for a step-by-step walkthrough. Note that GitHub does not render base64 embedded images, so you'll need a separate markdown viewer.
py.ens.Operation– executes Python code and sends back the results.
ml.engine.TrainProcess– trains a new prediction model.
ml.engine.PredictService– a service that receives information from engine sensors and sends it to ml.engine.PredictProcess to predict engine state. At the moment it is disabled (grey) and does not transfer data.
ml.engine.PredictProcess- uses the ML model to predict engine state.
ml.engine.CheckService– regularly checks the accuracy of the ML model. If the prediction error rate is above the threshold, the service sends a request to the
ml.engine.TrainProcessto update the model.
ml.engine.InitService– sends a request to the
ml.engine.TrainProcessto train the initial model at production start.
Predict Service and
Check Service to see model be automatically retrained when prediction quality drops.