No permission to save my model in work/storage

Hello Maxime,

The error message you're seeing is related to file permissions in your local environment, not with QuantConnect's platform. The code you're using is intended for use within the QuantConnect environment, where the ObjectStore is a special storage location provided by QuantConnect.

When you call qb.ObjectStore.GetFilePath(model_key), it returns a file path where the model will be stored in the QuantConnect environment. However, when you're running this code on your local machine, it seems like the path returned by GetFilePath is not writable in your local environment, hence the PermissionError.

If you're trying to save your model locally, you can simply specify a local file path where you have write permissions. For example:

file_name = "/path/where/you/have/write/permissions/model.pkl"
joblib.dump(model, file_name)

If you're trying to use QuantConnect's ObjectStore to save your model for use in a QuantConnect algorithm, you should run your code in the QuantConnect environment, either in a backtest or a research notebook.

Please refer to the QuantConnect Documentation for more details on how to use the ObjectStore in QuantConnect.

I hope this helps! Let me know if you have any other questions.

Hello and thank you for your answer,

However, I am not coding on my local environement but on QuantConnect's platform - and I am getting this error online, on the Jupyter notebook on the platform.

Kind regards,

Maxime

Hello Maxime,

I apologize for misunderstanding your initial question. If you're working in the QuantConnect research environment and encountering this issue, it seems like there might be a problem with the way the model is being saved.

In the QuantConnect environment, you should use the SaveBytes and ReadBytes methods of the ObjectStore to save and load your model. Here's an example of how you can do this with a scikit-learn model:

from sklearn import svm
from sklearn import datasets
import joblib
import io

clf = svm.SVC()
X, y= datasets.load_iris(return_X_y=True)
clf.fit(X, y)

# Save the model to a byte stream
model_byte_stream = io.BytesIO()
joblib.dump(clf, model_byte_stream)
model_byte_stream.seek(0)

# Save the byte stream to the ObjectStore
model_key = "model"
qb.ObjectStore.SaveBytes(model_key, model_byte_stream.read())

And here's how you can load the model back from the ObjectStore:

# Load the byte stream from the ObjectStore
model_byte_stream = io.BytesIO(qb.ObjectStore.ReadBytes(model_key))

# Load the model from the byte stream
clf = joblib.load(model_byte_stream)

Please give this a try and let me know if you're still encountering issues.

Hello,

The issue remains, it seems that I can save and load the model remaining in the same Jupyter notebook, but as soon as I want save the model from the Jupyter notebook and load it in the .py file (same project), the model doesn't exist.

Hello Maxime,

Based on the information provided, it seems you're able to successfully save the model in the Jupyter notebook environment using QuantConnect's ObjectStore but are encountering issues when trying to access it from a .py file within the same project. This scenario typically involves ensuring that the model is correctly saved and that the .py file is correctly attempting to access the model.

While the specific code to save and load models using ObjectStore in QuantConnect's environment was not directly provided in the extracts, the general approach to troubleshoot this issue involves a few steps:

1. Ensure Model is Saved Correctly: In your Jupyter notebook, after saving the model to the ObjectStore, verify that it exists by checking the return value of the SaveBytes or Save method (depending on how you're saving the model). QuantConnect's ObjectStore should return True if the operation was successful.

2. Check ObjectStore Availability: In your .py file, before attempting to load the model, check if the model key exists in the ObjectStore. You can use the ContainsKey method for this purpose. This step is crucial to ensure that the model was saved correctly and is accessible.

3. Load the Model Correctly: When loading the model in your .py file, ensure you're using the correct method (ReadBytes or Read depending on how the model was saved) and that you're correctly handling the data to reconstruct the model object.

Since the specific code snippets for saving and loading models using ObjectStore were not provided in the extracts, I recommend referring to the QuantConnect documentation on Storing Data with ObjectStore for the exact methods and usage.

If you're following these steps and still facing issues, it might be helpful to review any error messages or logs for more insights. Additionally, considering reaching out on the QuantConnect forum or support for more targeted assistance, especially if there might be a platform-specific issue at play.

I hope this guidance helps you troubleshoot the issue. If you have more specific error messages or code snippets, feel free to share them for more detailed assistance.