Is your feature request related to a problem? Please describe.
With all other parameters kept equal, high idark (i.e., less sensitive detector) results in more planet detections and fewer planet characterizations than a lower idark. A lower dark current probably permits time-consuming planet characterizations that the higher dark current is incapable of even attempting. Also, the lower dark current tried and failed more detections.
So, I ran detection-only simulations by setting n_det_min to 2 and max_successful_dets and nVisitsMax to 1, but the result was actually worse. Higher dark current results in even more planet detections, even without characterization added to the mix. It appears that the key issue is that lower dark current results in more failed detections.
My conclusion is that characterizations are expensive, failed attempts even more so. A better detector attempts more difficult targets and often fails. While failed attempts still return useful information, a worse detector paradoxically appears to yield more planet detections. I have not found a way to do an apples-to-apples comparison for dark current.
Describe the solution you'd like
Some options I have thought about:
-
An ‘omniscient’ detection-only mode where we know where the planets are and which detections will fail such that they are removed from the target list. The focus is only on the successful detections for each simulation.
-
An ‘omniscient’ characterization-only mode to reduce the uncertain effect of mixing in detections.
-
Order targets based on least to most integration time as opposed to completeness.
Additional context
Is your feature request related to a problem? Please describe.
With all other parameters kept equal, high
idark(i.e., less sensitive detector) results in more planet detections and fewer planet characterizations than a loweridark. A lower dark current probably permits time-consuming planet characterizations that the higher dark current is incapable of even attempting. Also, the lower dark current tried and failed more detections.So, I ran detection-only simulations by setting
n_det_minto 2 andmax_successful_detsandnVisitsMaxto 1, but the result was actually worse. Higher dark current results in even more planet detections, even without characterization added to the mix. It appears that the key issue is that lower dark current results in more failed detections.My conclusion is that characterizations are expensive, failed attempts even more so. A better detector attempts more difficult targets and often fails. While failed attempts still return useful information, a worse detector paradoxically appears to yield more planet detections. I have not found a way to do an apples-to-apples comparison for dark current.
Describe the solution you'd like
Some options I have thought about:
An ‘omniscient’ detection-only mode where we know where the planets are and which detections will fail such that they are removed from the target list. The focus is only on the successful detections for each simulation.
An ‘omniscient’ characterization-only mode to reduce the uncertain effect of mixing in detections.
Order targets based on least to most integration time as opposed to completeness.
Additional context