I run 3DVA with symmetry expanded particles and displayed the result with the cluster mode (6 clusters). I selected 5 of clusters to run ‘remove duplicates’ but the particles were greater than that should be. I did change the minimum separation distance to 0 and shift key to ‘none’. If I use all of particles in all clusters to run the job, there would be no problem. I don’t know why this happened. I am using CS4.6.2, BTW. Thank you so much for your kind suggestions in advance!
If you set minimum distance to zero, it defines particles which are 0.1Å dislocated as unique. Set Minimum separation distance to, e.g.; 50% of your particle diameter and see what effect that has.
I have the same problem and even setting a bigger minimum separation the problem is still there. I mean, it removes more particles, but still it retains more that it should be. However, I think the job should remove only the overlapping particles, therefore setting the distance to 0 should be the right way.
Hi @plvdac1! In most cases the minimum separation distance should be set to the diameter of your particle (or close to that value). We talk about this a bit in this section of the new 2D Classification guide page.
Essentially, what @rbs_sci says above is exactly right. Remember that this separation istance is the distance between the center of the particles. If you set the separation distance to 0 (or some very small value), even particles whose centers differ by a single pixel will be counted as unique. In practice, two unique particles will never be this close to each other – they’d be overlapping almost entirely!
Since even particles which are just touching touching (i.e., a separation distance of 1 particle diameter) are often too close to be useful, 1 particle diameter is a good starting place for duplicate detection.