Closed TeriForey closed 4 years ago
Hi Teri,
Sorry about this - I had never come across this error before. I have removed the "pause" statement from the Fortran code so it should continue to run now. I have created a new version of the Python package on PyPI: version 0.2.6.
I had a play with it and I can see why the warning is there in the T01 model - the trace goes pretty crazy when the starting point is far back in the tail. I started at (-25.0,0.0,0.0) to test it and it traces to distances with exponents of > 20. Interestingly, the same starting position used with the T96 model produces a realistic trace. It might be worth switching models in your code at X~-15 Re if you require traces from those positions.
Cheers!
Thanks Matt,
I'm using the T96 model by default, but I wanted to let users change the model if they wanted. Is it only the T01 model that returns this kind of error? If so I can add my own error message to warn them if the X~-15 Re.
Not a problem at all - yeah, just the T01 model as far as I am aware. I'm not sure at what point the other models become highly unrealistic, but my grep
search didn't find any more pauses in any of code for the other models.
Hi Matt,
I'm running PyGeopack within a web server and it's working really well except that occasionally some of the values are generating an error when the user changes the model choice. Unfortunately instead of being a raised exception that I can capture and feedback to the user - it's occurring within the Fortran code and just kills my server.
The kind of error I'm getting is:
Obviously I can't have the server crashing when a user makes a bad choice, is there any way these kind of errors could be caught within your library so it protects my server?
Thanks,
Teri