Save Model from GUI and use it in Python

[jonathanhoss]

Hello,
I want to save a model which is trained in the GUI and load it afterwards in Python to deploy it as a microservice.
I saved the model as .pkcls File and loaded it into Python. When I try to run the predict function, it gets an ValueError.
How can I access the preprocessing processes that is done in the GUI?
And How do I integrate the preprocessing in my python-script?

import pickle
import Orange
import numpy as np

with open('naive-bayers.pkcls', 'rb') as f:
    mp = pickle.load(f)
    
print(mp)

arguments = np.array([521, 24, 'female', 2, 'own', 'little','little', 3190, 18, 'radio/TV']).reshape(1, -1)

print(arguments)
>>> [['521' '24' 'female' '2' 'own' 'little' 'little' '3190' '18' 'radio/TV']]

print(mp.predict(arguments))
>>> ValueError: Unable to convert array of bytes/strings into decimal numbers with dtype='numeric'

[markotoplak]

You have to put the data into a learner-compatible Orange.data.Table (basically, 3 numpy arrays + a Domain describing your columns) to use Orange models. Here are some docs on creating tables from code:

• https://orange3.readthedocs.io/projects/orange-data-mining-library/en/latest/reference/data.table.html