HWP angle correction

[ykyohei]

I have tried to organize what I was thinking about how to correct the HWP Angle.

Now we have no simple fix like below that anyone can use.

hwp_angle_true = sign_convention * (hwp_angle_encoder + offset)

There are two reasons for this. One major reason and another minor reason.

• The sign_convention is not obvious value for SO hwp encoder, and we found that we have mis-estimation of sign_convention in some observations. Therefore, we are preparing multiple estimation methods for sign_convension. https://simonsobs.atlassian.net/wiki/spaces/PRO/pages/185434251/CHWP+main+page#Rotation-direction-definition
• The offset of hwp fast axis can be different for different band (and also detectors), because SO HWP is achromatic.

Therefore, the actual hwp angle correction equation will be like this.

hwp_angle_true = sign_absolute_satX * sign_relative_between_observations * (hwp_angle_encoder + offset_absolute + offset_relative_between_band/det + offset_relative_between_observations_due_to_hwp_sag)

We have two hwp encoders, and they have different offset_absolute.
offset_relative_between_observations_due_to_hwp_sag depends mainly on elevation and cooldown.

DIscussion about pipeline implementation

My original plan was that make_hwp_solution.py will calculate hwp_angle_true and write it as hwp_angle metadata.
But, I think making a following correction module will simplify and clarify the correction of hwp angle.

• hwp_angle metadata will have hwp_angle_encoder, multiple sign_relative_between_observations , and offset_relative_between_observations_due_to_hwp_sag .
• Separate metadata will have offset_relative_between_band/det from wire grid calibration etc.
• Config file will have constant correction factors: sign_absolute_satX and offset_absolute
• We make a hwp loading module that solve the consistent and naturally defined hwp_angle_true by the equation above.
  • We probably need to consider more about how and when we correct offset_relative_between_band/det.