Every so often the session state for the Advanced Data Analysis model is reset.
When this happens all of the temporary variables created during the season are deleted and any files in your sandbox are removed.
There is no way to recover from this.
All you can do is re-run the commands in a new session to get back to where you were.
I’ll share my end of session steps which may help you in the future.
The first thing I like to do is, at the end of a session, ask the model to generate a python file with all of the code necessary to get back to this point.
This way if I want or need to make any changes to the code, I can manually.
I then ask it to store all session variables to a .pkl
file to be reloaded later.
Lastly, I ask it to write a clear and concise summary of what the purpose of this session and what we’ve done so far into a text file, I ask it to include anything we’ve tried that failed and why it didn’t work so if we need to start over we don’t end up going down the wrong path again.
Then I ask it to wrap up all of the files in my sandbox into a archive file which I can download.
When I need to restore a session, I can simply upload the archive file and one of,
- Load everything back into the workspace from the
.pkl
file using globals
- Run the
.py
script to regenerate everything
- Ask it to read the
summary.txt
file to understand the task and work again from first principles.
It might seem silly and redundant to have all three available, but depending on what I want to do moving forward it can be helpful to have all the options available to me.
Again, none of this will help you now but going forward—if you’re consistent about doing it—I think it will greatly alleviate a great many issues.