Hi all,
I would like to ask you for some help with an error I got while running Topaz Train (beta):
Andrea
Hi Andrea,
It looks like Topaz is not in the path for Cryosparc. Is Topaz installed? Topaz must be installed separately from Cryosparc.
Best,
-Alex
Hi Alex,
thank you for your prompt reply and to develop Topaz.
Yes, I installed Topaz separately from cryosparc. With the terminal command “which topaz” I got the path (I can’t explain why it has been installed inside eman2 folder, by the way. Could be the error relies on the installation folder (eman2)?
Any suggestion is more than welcome.
Thank you,
Andrea
Hi Andrea,
Was Cryosparc launched from the same terminal that has Topaz activated and in the environment?
Best,
-Alex
Hi Alex,
yes, the terminal is the same for both cryoSPARC and Topaz, and Topaz has been activated as explain both in cryosparc guide and your github page.
Even Topaz Denoise give me the same error.
Did I install Topaz in the wrong way?
Best,
Andrea
Hi Andrea,
It sounds like you did everything correctly as explained in the Cryosparc documentation. Perhaps the Cryosparc team can help you track down why exactly it isn’t connected properly.
Best,
-Alex
Hi Alex,
thank you for your support.
Apparently, now Topaz is working. I had to downgrade to version 0.2.3. Topaz Train (beta) worked, but Topaz Denoise (beta) didn’t, it gave me a " ValueError: min() arg is an empty sequence". I’m working on that.
Another thing, I cannot find Topaz Inference to observe the particles picked with Topaz Train. Is it present only in the last Topaz version?
Thank you again.
Best,
Andrea
Hi Andrea,
You can use Manual picker and Inspect picks jobs to look at Topaz picks.
Best,
-Alex
Hi Alex,
thank you again. Yes, I used Inspect picks and it works perfectly.
I would like to ask you some details about Topaz Train, I tried to check on the web but I couldn’t find anything. Are these UserWarning normal or there is some error?
Hi Andrea,
I’ve never seen that before. Does the job complete?
Best,
-Alex
Hi Alex,
both Topaz Train and Topaz Cross Validation gave me the same warning, but then both of them finished without any problem.
I noticed that Topaz extract, after Topaz cross validation, recognized the same particle multiple times. I think that I have to remove duplicate.
Is it normal?
Thank you.
Best,
Andrea
Hi Andrea,
Increase the extraction radius in Topaz Extract until you get one pick per particle.
Best,
-Alex
Hi Alex,
it works, thank you!
I would like to ask you about an error I got with Topaz Denoise:
Andrea
Hi Andrea,
Honestly, I haven’t tried training a Topaz Denoise model in Cryosparc. It looks like you ran into a Cryosparc error, not a Topaz error.
Is there a reason you want to train a new model instead of using the pre-trained models?
Best,
-Alex
Hi @alexjamesnoble,
Sorry to loop back to the original issue, but it seems I’m having a similar issue with “command “/home/exx/anaconda3/envs/topaz/bin/topaz --version” did not produce valid output:”, despite having 0.2.3 installed. If i just execute > /home/exx/anaconda3/envs/topaz/bin/topaz --version from this terminal it seems to provide the version just fine. Any suggestions here? For reference I’m on CS version 3.1.0
Thanks,
Nathanael
Hi Alex,
sorry to come to you so late.
No, there isn’t any reason, I just thought that it was the correct way to use Topaz with new micrographs and particles.
Good to know that the pre-trained model works as well!
Sorry for the misunderstanding.
Thank you for your support.
Best regards,
Andrea
