cannot convert float NaN to integer
pent1162 opened this issue · comments
Traceback (most recent call last):
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\gui_qt\job_runners\worker.py", line 60, in run
result = self.fn(*self.args, **self.kwargs)
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\gui_qt\job_runners\denoise_job_runner.py", line 67, in start_func
denoised = self.denoiser.denoise(
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\restoration\denoise\classic.py", line 349, in denoise
response = self.it.translate(
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\it\base.py", line 393, in translate
tilling_strategy, margins = self._get_tilling_strategy_and_margins(
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\it\base.py", line 569, in _get_tilling_strategy_and_margins
spatiotemp_tilling_strategy = tuple(
File "C:\Users\xxxxxx\Miniconda3\envs\aydin_env\lib\site-packages\aydin\it\base.py", line 570, in
max(1, min(rest_split_factor, int(math.ceil(split_per_dim * s))))
ValueError: cannot convert float NaN to integer
System:
Intel Xeon Gold 6256 CPU 3.6Ghz
RAM 1TB
Windowns Server 2016 ver1607
could you share the image file you were using with me @pent1162 ? I would like to reproduce the error to fix it correctly.
Hi @pent1162 ,
I realized this is a two-channel image. Trying to denoise a multi-channel image as is won't be my suggestion(models usually have hard time to benefit from correlations across different imaging channels). So I downloaded your image and split its channel. I did it for the first channel, here you can see the result:
Also please let me know what is your training crop selection. If you like to continue on this discussion in private, feel free to send me an email.
Happy new year @pent1162 ! I wanted to check if you were able to run Aydin on your image with your computer? Please let me know if there is anything I can help with.
Happy new year.
It worked well and thanks for your help.
Thank you for letting me know. I will close this issue for now but feel free to reach out to me if there is anything else I can help with!