create function posts values to model dataset
Dears, im clearly new to Dataiku,
i have built a model based on 2 sources of datasets:
1- one directly from sql (which includes most of the columns)
2- and the second uploaded excel file with 2 external feautes filled manually.
Then i joined them to create my dataset (dataframe) and build the flow and model.
Now to use the model for prediction; i need to ensure the the model recives all the values for all its columns,
the user must input 2 features + and the file number(to get from DB), AND my API should pushall the data to the model in order it can predict the answer.
where should i build this function that recives the 2 input values from the user ? in the API deployer/service / node /endpoint?
Best Answer
-
Hi,
Check out this tutorial on creating a project application that allows a user to attach an excel file which will then pass through the flow and then use the model to make predictions (note that it would be a slight change to have the user input a DB file name instead of drag/drop a static csv): https://knowledge.dataiku.com/latest/courses/o16n/dataiku-applications/create-dataiku-app.html
Best,
Pat
