MembraneCurvature very slow with low number of frames

[cyril-schroeder-ox]

Hi all,

I noticed that when calling MembraneCurvature on only a single frame, the performance becomes very bad compared to calling it over a couple of frames (~50 sec compared to <1 sec). Is there an algorithmic reason for this?

Thanks a lot in advance!

[cyril-schroeder-ox]

In case this is useful to somebody else, while the above behaviour was indeed an issue, my other problem of obtaining NaN values came from having proteins inserted into the membrane.