Appion

Hard to say, confidence is a good measure when comparing different packages and last I checked it was the default filter in makestack.

If I am tracing the function right, makestack uses apCtf.ctfdb.getBestCtfValue function without sortType specified. Therefore, it is using res80 as primary scoring as default. Only if res80 is not available, it will use calculateConfidence function that takes the maximum of 30/10, 5_peaks, and the package defined confidence. How about displaying them as one item ?

80% of phase plate data which in general has defocus of -0.5 um could not generate apCtf.ctfdisplay analysis results. I need to show package resolution while fitting all of the displays in one line.

Yes, but I think it uses the 80% confidence before entering the loop to outright reject images. Check the pre-loop functions.

I hope to take a look at fixing my stuff for phase plates when I get time again.

Fixing ctfdisplay analysis would be a different issue. I, for one, don't use that cutoff in makestack any more because they have little predictive power for high resolution data analysis.

The 80% confidence cut off for rejecting image also uses calculateConfidenceScore. Therefore, I am going to display one value based on that algorithm. Call it conf (Appion) unless there is objection.

Seemms like a good plan to me - I also find these conf values very confusing.

Thanks Anchi, I think your plan is good.