Closed ghost closed 5 years ago
Could you please provide a minimum example?
Possibly a duplicate of gh-3521
Yes the issue posted in gh-3521 is similar with what I am facing. But I didn't get anything where it shows the issue is resolved. Can anyone direct me to the right post or link.
It is not resolved yet, unfortunately. 😞 We will discuss this at Friday's meeting.
I am writing a python script so that the pickle file generated by the orange save data widget is used in the python script. So can anyone tell me how to convert a pandas data frame into orange data table so that it can be consumed in the pickle file in the same format as it does in orange?
You need to construct an ordinary numpy array and a corresponding Domain, and pass them to Orange.data.Table.from_numpy.
Orange version
Orange version - 3.22.0
Expected behavior
I am using orange, to save the model trained on orange canvas to a pickle file using the save model widget. The model gives the correct predictions when loading the test and train data during the model development. But when I save the model using the save model widget in pickle file and use it on a new canvas using load model widget, it gives the wrong output. Basically it only gives out one class for all the test data passed for prediction when using load model.