ctf bfactor

[geoffwoollard]

RE https://github.com/harshit-gupta-cor/simSPI/blob/linear_simulator/simSPI/linear_simulator/modules.py#L235

How exactly does the code below make a good bfactor?

• valueNyquist is 0.2 here https://github.com/harshit-gupta-cor/simSPI/blob/linear_simulator/simSPI/linear_simulator/configs/betagal_starfile.cfg#L40

• the minus log (0.2) goes to some positive number: 1.61

• decay is 1.612^2pixel_size = 6.44*pixel_size

• what is the rhyme or reason?

        if self.config.bfactor == 0:
            decay = (
                np.sqrt(-np.log(self.config.valueNyquist))
                * 2.0
                * self.config.pixel_size
            )
            envelope = torch.exp(-self.frequency ** 2 * decay ** 2 * self.r2)

[geoffwoollard]

see initial discussion in this PR https://github.com/compSPI/simSPI/pull/20#discussion_r759614238

[geoffwoollard]

by "default", I mean that when b=0, we enter this control flow