Classifier Fetch Images Bug

We have been successful at fetching images for all conditions (random, positive, negative, etc.) when we try to select cells from the entire “experiment”, but when we change from “experiment” to a specific image, we can only fetch images when the condition is set at random, for example if we want to fetch 20 positive cells from specific image, the following message pops up in a window.

An error occurred in the program:
OperationalError: Expression tree is too large (maximum depth 1000)

Traceback (most recent call last):
File “ClassifierGUI.pyc”, line 559, in OnFetch
File “MulticlassSQL.pyc”, line 108, in FilterObjectsFromClassN
File “DBConnect.pyc”, line 389, in execute

We have tried only pulling one or two images and we get the same error.

I have also attached the text that is displayed within the console when the error occurs.

console (11.5 KB)

Are you using SQLite?

It does appear to be an SQLite error message, and I can see the cause. I will add a fix to SVN.

In the meantime, you might be able to work around this by scoring the image you wish to select from, and dragging cells directly from that image.

Thanks for the help and working on the fix.


Are you using the compiled version of CPA, or running a source-based installation via SVN?

Just the complied version for now. Running the source-based installation I will leave for my programmers when they have a chance or when you guys release a new version, whichever happens first.