Table 1.
Execution time per light curve for simulations containing either 1000 data points (Data set A) or 3840 data points (Data set B) for the transit of a star by a planet with p ≈ 0.1.
| Algorithm | Limb darkening | Processor | Execution | Notes |
|---|---|---|---|---|
| time (μs) | ||||
| 5@l Data set A | ||||
| qpower2, pycheops | power-2 | 2 GHz CPU | 113 | Optimised using numba |
| qpower2, pycheops | power-2 | 2 GHz CPU | 5550 | No optimisation |
| qpower2, Fig. A.1 | power-2 | 2 GHz CPU | 570 | |
| batman | quadratic | 2 GHz CPU | 169 | |
| batman | power-2 | 2 GHz CPU | 2380 | maxerr = 1 (default) |
| batman | power-2 | 2 GHz CPU | 247 | maxerr = 70 |
| Data set B | ||||
| qpower2, C/OpenMP | power-2 | 4.8 GHz CPU | 357 | |
| qpower2, C/OpenMP | power-2 | 4.8 GHz CPU ×8 | 112 | |
| qpower2, C/OpenMP | power-2 | GTX1080 GPU | 13.2 | Return array of models |
| qpower2, C/OpenMP | power-2 | GTX1080 GPU | 2.5 | Return  values only |
Notes.
The notes for batman give the value of the max_err option that is used to set the number of integration steps such that the maximum error due to numerical noise does not exceed the value given in ppm.
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.