Encountering ModuleNotFoundErrror

[MallikaGakhar]

Hello,

I am trying to run an object through the webpage. It's a very small dataset but I keep encountering ModuleNotFound error for starcat library. I would appreciate if you could help me with the same.

[dylkot]

Hi, thanks for letting me know about this. Yes there does seem to be an issue with the website. I'll work on fixing it. For now, you can try using the python or command line API. I'll update this when I fix it. Thanks!

[dylkot]

Hi, this should now be fixed. Thanks!

[carolinturner]

Hi, I'm encountering the same problem. Do you know whether it takes some time for the fix to take effect? Thanks!

[dylkot]

Odd, initially I was getting the same error and I tracked it down to an updated version of AnnData being used with scanpy that wasn't compatible with the website. But after I pinned an older version of AnnData, I am not getting it any more. Can you copy the whole text that is showing up below the "Run starCAT" box?

[carolinturner]

Thanks for looking into this.

Here is the full error message:

Launching starCAT...
An error occurred during the Python execution: Traceback (most recent call last):
File "/lib/python312.zip/_pyodide/_base.py", line 596, in eval_code_async
await CodeRunner(
File "/lib/python312.zip/_pyodide/_base.py", line 412, in run_async
await coroutine
File "", line 2, in
ModuleNotFoundError: No module named 'starcat'

[dylkot]

thanks and which web browser are you using? I just ran it with two datasets on Chrome and Edge and both worked on both.

And can you summarize the text that comes up initially when you refresh the page that then dissapears? It should be:

Loading pyodide...
Loading micropip...
Loading anndata (pure python version)...
Loading starcatpy...

If you want, you can send me your data to test and I can take a look although the issue seems to be more with setting up the analysis rather than with the data itself.

[carolinturner]

I am using Edge and after refreshing the page, it now works.

It first shows the text that you mention:
Loading pyodide...
Loading micropip...
Loading anndata (pure python version)...
Loading starcatpy...
And when I then re-upload my data, it works.

Thanks for being so responsive!

[dylkot]

Great, thanks for confirming!

[MallikaGakhar]

Hey, sorry to bother you but I encountered "query input contains negative values. Must be non-negative"

Loading reference from existing cache file for reference TCAT.V1
An error occurred during the Python execution: Traceback (most recent call last):
File "/lib/python312.zip/_pyodide/_base.py", line 596, in eval_code_async
await CodeRunner(
File "/lib/python312.zip/_pyodide/_base.py", line 412, in run_async
await coroutine
File "", line 34, in
File "/lib/python3.12/site-packages/starcat/starcat.py", line 213, in fit_transform
query = self.prepare_query(query)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lib/python3.12/site-packages/starcat/starcat.py", line 241, in prepare_query
raise Exception("""query input contains negative values. Must be non-negative""")
Exception: query input contains negative values. Must be non-negative

But if I run the object using the reference statCAT script, it extracts counts correctly. Is there a difference between how counts are being extracted/processed here on the wwebpage and use the refernce starCAT ipynb file?

[dylkot]

Interesting. No there shouldn't be a difference in the starCAT commands themselves. It is possible the difference could be due to the version of AnnData or Scanpy which could lead to some subtle differences. This would be pretty useful for me to check. Could you check the version of AnnData, scanpy, and starcat you are running. If you want to email me the AnnData file, I could look into it as well.

[MallikaGakhar]

Hey, I did email you that day and not sure if it reached you but I tried using the tool again today and this time the webpage just crashed even though its the same dataset.