Closed maximilianh closed 7 years ago
Because of the way the parser works, if there's an issue with the predicate or the object, it will try and just add the subject (since it lexes from left to right) so I've added in a warning message. Like it said, this just means it's worth checking.
The Selventa large corpus makes a lot of entries for the existence of complexes that aren't part of a relationship, so it makes these warnings. Maybe it's a good idea to add an option to suppress these (if you're sure that your BEL is coded properly)
OK, thanks!
I successfully ran through, I now have the CSV file. This is an awesome converter!
On Fri, Jun 2, 2017 at 2:42 AM, Charles Tapley Hoyt < notifications@github.com> wrote:
Because of the way the parser works, if there's an issue with the predicate or the object, it will try and just add the subject (since it lexes from left to right) so I've added in a warning message. Like it said, this just means it's worth checking.
The Selventa large corpus makes a lot of entries for the existence of complexes that aren't part of a relationship, so it makes these warnings. Maybe it's a good idea to add an option to suppress these (if you're sure that your BEL is coded properly)
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/pybel/pybel/issues/200#issuecomment-305740691, or mute the thread https://github.com/notifications/unsubscribe-auth/AAS-TZckf0MI4vRPJejv2rCTtsJkQ1hdks5r_9kYgaJpZM4NtmmA .
You may want to upload the .cx result of the large corpus onto ndex. But please also add a link to the .csv file, the .cx file format is terribly hard to parse and understand. .csv is a lot better.
On Fri, Jun 2, 2017 at 2:44 AM, Maximilian Haeussler maximilianh@gmail.com wrote:
OK, thanks!
I successfully ran through, I now have the CSV file. This is an awesome converter!
On Fri, Jun 2, 2017 at 2:42 AM, Charles Tapley Hoyt < notifications@github.com> wrote:
Because of the way the parser works, if there's an issue with the predicate or the object, it will try and just add the subject (since it lexes from left to right) so I've added in a warning message. Like it said, this just means it's worth checking.
The Selventa large corpus makes a lot of entries for the existence of complexes that aren't part of a relationship, so it makes these warnings. Maybe it's a good idea to add an option to suppress these (if you're sure that your BEL is coded properly)
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/pybel/pybel/issues/200#issuecomment-305740691, or mute the thread https://github.com/notifications/unsubscribe-auth/AAS-TZckf0MI4vRPJejv2rCTtsJkQ1hdks5r_9kYgaJpZM4NtmmA .
What does this mean:
Is it normal that I'm getting tens of thousands of warning with the "official" large corpus file?