Table C.4
Relative log Bayesian evidences, Δ ln 
, as computed by PolyChord for non-GP and multidimensional (MD) GP models including different numbers of Keplerian terms, both with and without the inclusion of HIRES data.
| Non-GP modelling of RVs only | MD-GP modelling of RVs, BIS,  |
|||
|---|---|---|---|---|
| Keplerians | (i) HARPS-N only | (ii) HARPS-N + HIRES | (iii) HARPS-N only | (iv) HARPS-N + HIRES |
| – | −75.98 ± 0.16 | −77.13 ± 0.21 | −76.24 ± 0.14 | −73.91 ± 0.14 |
| b | −42.2 ± 0.13 | −38.08 ± 0.14 | −38.19 ± 0.21 | −33.51 ± 0.18 |
| b, c | −7.32 ± 0.12 | −6.14 ± 0.16 | −9.12 ± 0.12 | −6.5 ± 0.13 |
| b, c, d | −0.94 ± 0.41 | −1.55 ± 0.46 | −1.6 ± 0.53 | −1.07 ± 0.18 |
| b, c, d, e | 0 ± 0.32 | 0 ± 0.13 | 0 ± 0.38 | 0 ± 0.13 |
| c | −49.79 ± 0.12 | −56.51 ± 0.13 | −51.92 ± 0.12 | −55.64 ± 0.13 |
| d | −59.24 ± 0.18 | −63.01 ± 1.49 | −60.15 ± 0.95 | −63.25 ± 0.30 |
| d, e | −54.87 ± 1.50 | −60.61 ± 1.33 | −58.82 ± 0.21 | −51.23 ± 4.81 |
Notes. Uncertainties correspond to the standard error on the mean evidence across three PolyChord runs, or to the highest individual run error provided by PolyChord (the greater of the two). For ease of interpretation, the highest evidence in a given column is normalized to zero; all other evidences in that column are defined relative to that zero point. Note that evidences cannot be compared meaningfully across setups (i)–(iv), as the data being modeled are different in each case. In each setup, a three-planet model is strongly favored, with weak but insufficient statistical evidence in support of a four-planet model.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.