[Bug]: Cannot find object in prediction using various models

[vladimirsim]

JASP Version

0.19.1

Commit ID

No response

JASP Module

Machine Learning

What analysis are you seeing the problem on?

Prediction

What OS are you seeing the problem on?

Windows 11

Bug Description

The prediction table based on a trained KNN model did not load.

Expected Behaviour

The prediction table should load.

Steps to Reproduce

1. Train a KNN model on a training dataset and save it
2. Open Machine Learning> Prediction>Prediction
3. Load a test or prediction dataset (different from the training dataset) and load the trained model
4. Load the features from the trained model. The prediction table fails to load.
   I am attaching a printscreen:
   

This is the trained KNN model:
KNNTrainedModelVladimirsim.zip

...

Log (if any)

JASP 2024-11-28 10_52_36 Desktop vladimirsim.zip

More Debug Information

This is the error message which I got while trying to predict values using a trained KNN model:
This analysis terminated unexpectedly.

Error in eval(predvars, data, env): object 'Good' not found

Stack trace
jaspStateR$new(object = object, dependencies = dependencies)

initialize(...)

.mlPredictionGetPredictions(model, dataset)

.mlPredictionGetPredictions.kknn(model, dataset)

kknn:::predict.train.kknn(model[['predictive']], dataset)

kknn(formula(terms(object)), object$data, newdata, k = object$best.parameters$k, kernel = object$best.parameters$kernel, distance = object$distance)

model.frame(formula, data = train)

model.frame.default(formula, data = train)

eval(predvars, data, env)

eval(predvars, data, env)

To receive assistance with this problem, please report the message above at: https://jasp-stats.org/bug-reports

Final Checklist

• I have included a screenshot showcasing the issue, if possible.
• I have included a JASP file (zipped) or data file that causes the crash/bug, if applicable.
• I have accurately described the bug, and steps to reproduce it.

[koenderks]

jasp-stats/jaspMachineLearning#393 should have fixed this for 0.19.2, but I can't know for sure if I don't have the data. You might test this fix yourself using the latest development version from https://static.jasp-stats.org/Nightlies/

[vladimirsim]

Thank you, @koenderks , but this is exactly the version which I am using and which produces the errors. I am using 84d54b934fa27731bb9eec44a4aa5f7ab0744dfd, which I downloaded and installed yesterday. The only difference is that I am using the HEAD instead of the 'development' version, but when I downloaded the 'development' file, I see that it is the same size (923512 KB) as the HEAD version. So I assume that they are the same. Or am I wrong? Should I still install the 'development' version and give it a try?

[koenderks]

The bug fix was only incorporated into the development version yesterday (23 hours ago, 84d54b934fa27731bb9eec44a4aa5f7ab0744dfd is from 2 days ago), so I would assume that only the latest nightly from 27-11-24 (in the evening) might have this fix in it. Could you try the latest version?

[vladimirsim]

Yes, it worked! The 'explain predictions' table worked for all methods, too!
A huge thank you, dear @koenderks for your invaluable and lightning-fast support!

[koenderks]

Good to hear! I'm closing this issue then