I notice that matroid code does not make any use of @cached decorator. Perhaps it could simplify parts of code it it was used.
On Thursday, March 17, 2016 at 7:54:53 AM UTC, Rik Schepens wrote: > > In my first steps to improve matroid visualization as part of my Bachelor > thesis, I wanted to use the cache option as is implemented by Jayant Apte > to immediately visualize some standard (named) matroids. For more > information see ticket #20175. > My current solution writes the information directly on _cached_info, as is > done multiple times in other pieces of code. This _cached_info is just used > to store this positioning information. My thesis supervisor (Rudi > Pendvingh) mentioned it might be better to use a method instead of saving > it directly. This method is already in place and is called _fix_positions. > Using this method works just fine if we use it just for adding this > information for the matroids in the catalog. Though replacing current > (direct saving) code in the plot method within this method gives some > errors I wasn't able to identify. (see attachment) > > I would love to know some of your opinions on whether to saving directly > or using the current method is the preferable way to implement these > standard visualizations. And I was also wondering if this method is the > preferable way, as I think it might be, then why do these errors appear > when this is used in other methods. > > Kind regards, > Rik > > > > -- --- You received this message because you are subscribed to the Google Groups "sage-matroid" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
