Problems with Satpy 0.37

38 views
Skip to first unread message

lobsiger...@gmail.com

unread,
Aug 11, 2022, 9:42:39 AM8/11/22
to pytroll
Dear developers,

After upgrading from SatPy 0.36 to 0.37 my MetopB/C scripts do not work as before. Below the stdio output of my MetopC script and the stderr output from 0.36 and 0.37 in two text files. I want to make the two attached images. It works for 0.36 on a PC not upgraded yet but fails in 0.37 (in GNU/Linux as well as Windows 10). Any ideas?

Regards,
Ernst

Satellite MetopC is expected to make an ideal NIG overhead pass in the center of area "isleofman" at 21:00 UTC.
Add worst case EUMETCast channel timeliness + 1 hour slack to above time for scheduling images from this area!
 1 --> /srv/E1B-EPS-10/2022/08/10/AVHR_xxx_1B_M03_20220810203703Z_20220810204003Z_N_O_20220810211140Z.bz2
 2 --> /srv/E1B-EPS-10/2022/08/10/AVHR_xxx_1B_M03_20220810204003Z_20220810204303Z_N_O_20220810211212Z.bz2
 3 --> /srv/E1B-EPS-10/2022/08/10/AVHR_xxx_1B_M03_20220810204303Z_20220810204603Z_N_O_20220810211243Z.bz2
Script: MetopC-area.py 20220810NIG POI: lat=54.2 lon=-4.5 ran=15.0
Maximum elevation occurs on orbit number=19503 at telmax=20:41 UTC




MetopC-20220810-NIG-2041-natural_with_night_fog-isleofman.jpg
MetopC-20220810-NIG-2041-ir108_3d-isleofman.jpg
ERR036.txt
ERR037.txt

Raspaud Martin

unread,
Aug 11, 2022, 10:03:02 AM8/11/22
to pyt...@googlegroups.com
Dear Ernst,

Thanks a lot for the feedback! it's very much appreciated.

The error is not something I recognize off the bat, so I just would
like to confirm with you that satpy is the only package you upgraded,
or if dask or xarray for example were upgraded as well?

In anycase, since we are probably going to need more information on
this, would you mind filing an issue on the satpy github page here
https://github.com/pytroll/satpy/issues and provide information as
instructed in the template? that will help us greatly track down the
problem.

Best regards,
Martin

lobsiger...@gmail.com

unread,
Aug 11, 2022, 10:33:56 AM8/11/22
to pytroll
Dear Martin,

unfortunately the only PC where I still have Satpy 0.36 has not been touched/used for a long time except for a rather recent update to 0.36.
So a lot of packages are quite different but have also worked in newer version until I upgraded the PC I usually work on to 0.37. As my scripts
are rather long as well I will try to nail the problem down with a minimum code and Metop data included that I can publish as wetransfer.
I'll then file an issue on github.

Best regards,
Ernst
conda_list_037.txt
conda_list_036.txt

lobsiger...@gmail.com

unread,
Aug 11, 2022, 2:12:09 PM8/11/22
to pytroll
Dear developers,

I made a short script including MetopC data and area definition 'isleofman' that reproduces the problem. It's 31MB zipped under this wetransfer link (valid 1 week):


Cheers,
Ernst

Panu Lahtinen

unread,
Aug 12, 2022, 1:29:33 AM8/12/22
to pyt...@googlegroups.com
On 11.8.2022 16.42, lobsiger...@gmail.com wrote:
> Dear developers,
>
> After upgrading from SatPy 0.36 to 0.37 my MetopB/C scripts do not work
> as before. Below the stdio output of my MetopC script and the stderr
> output from 0.36 and 0.37 in two text files. I want to make the two
> attached images. It works for 0.36 on a PC not upgraded yet but fails in
> 0.37 (in GNU/Linux as well as Windows 10). Any ideas?

Could you compare the versions of PIL between the installations? If they
differ, try downgrading the library in the new installation to the
previous version to see if that helps.


P

--
Finnish Meteorological Institute
Erik Palménin aukio 1, P.O.Box 503, FIN-00101 Helsinki, Finland
Room: 4B15a, Phone: +358 50 380 3261

lobsiger...@gmail.com

unread,
Aug 12, 2022, 2:33:30 AM8/12/22
to pytroll
Hi Panu,

in a post above I attached the full output of "conda list" on my working and non working pytroll environments.
Indeed with (GNU/Linux) satpy 0.36 I had pillow 9.1.1 while with (GNU/Linux) satpy 0.37 I have pillow 9.2.0.
I downgraded pillow 9.2.0 in my Windows 10 install to 9.1.1 with "conda install pillow=9.1.1". It did not help :-(.


Regards,
Ernst

lobsiger...@gmail.com

unread,
Aug 12, 2022, 2:59:46 AM8/12/22
to pytroll
Panu,

in my Windows environment I updated pillow to 9.2.0 again. Then I downgraded Satpy to 0.36. Screen capture below:

------------------

(pytroll) C:\EMCtools\LEODEV>conda install satpy=0.36
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: done

## Package Plan ##

  environment location: C:\Users\immpac\miniconda3\envs\pytroll

  added / updated specs:
    - satpy=0.36


The following packages will be DOWNGRADED:

  satpy                                 0.37.0-pyhd8ed1ab_0 --> 0.36.0-pyhd8ed1ab_0


Proceed ([y]/n)?

Preparing transaction: done
Verifying transaction: -
SafetyError: The package for satpy located at C:\Users\immpac\miniconda3\pkgs\satpy-0.36.0-pyhd8ed1ab_0
appears to be corrupted. The path 'site-packages/satpy/etc/composites/visir.yaml'
has an incorrect size.
  reported size: 14290 bytes
  actual size: 14276 bytes


done
Executing transaction: done

(pytroll) C:\EMCtools\LEODEV>

--------------------

Now my script works again as before.

Best regards,
Ernst

Raspaud Martin

unread,
Aug 12, 2022, 4:57:14 AM8/12/22
to pyt...@googlegroups.com
Dear Ernst,

I could run the script you have sent, and could reproduce the error.
To avoid confusion, I will continue reporting on the issue you created
in github (https://github.com/pytroll/satpy/issues/2173)

Best regards,
Martin
Reply all
Reply to author
Forward
0 new messages