A shared APF night correction leaves a 2.41 m/s component at 192.32 days in a fixed model of HD 219134

Can a velocity offset shared by all APF stars on each observing night explain the disputed 192.32-day component in HD 219134? We test public California Legacy Survey data with 62 reference stars and all 295 target nights. Each star has a separate quadratic trend. The target has sine and cosine terms at six fixed periods. We remove arbitrary shared night offsets by weighted contrasts within each night. The fitted 192.32-day semi-amplitude changes from 2.20 to 2.41 m/s. A residual sign bootstrap uses independent signs for each star and each 90-day block. Its conditional 95% interval for the corrected amplitude is 2.04 to 2.80 m/s. The interval for the amplitude change is -0.02 to 0.44 m/s, so the increase is not significant under this check. Reference and model checks retain the component. A reference-only calibration with its full error covariance gives the same result as the joint fit. The package includes derived nightly data, source hashes, code, figures, fit products, and numerical validation. Under the tested fixed-period model, a shared additive night term does not remove the component. This test does not establish a planet or exclude star-specific instrument errors, stellar activity, or other orbit models.

HD 219134 · radial velocity · Automated Planet Finder · night offsets · reference stars · public astronomical data · instrument errors · fixed-period model