Skip to content

Detection Tabulation

We make a tabulation of the detection process in which target stars or planets are found and a count of (hopefully) successful detections accumulates. There are two tables, each restricted to a specific target family:

  • All stars: Every star in the target list which was observed.

  • Promoted stars: a list of stars that have been promoted into a category where they are eligible for spectral characterization. The EXOSIMS scheduler controls promotion, and provides the list at the end of the simulation in the promoted_stars variable.

All quantities shown are ensemble averages of detection-related events across all DRM's in the ensemble - the ensemble mean.

All quantities after ± are the standard error of the mean. That is, they are the population standard deviation divided by the square root of ensemble size, so they will drop to zero as ensemble size grows. The ensemble size used for this scaling is shown in a note below the table. The intent is to show how accurate the shown average is.

Table Contents

Column naming. The final column is the earthlike planet class. Its heading follows the name given in the scenario's config-reduce.json, and is "Earth" when nothing is customized -- so a scenario that re-defines the class to, say, a Sub-Neptune population will show "Sub-Neptune" there instead. This page uses the default name throughout.

In all cases, consideration is restricted to a population P of stars (all observed stars, or promoted stars).

Every detection observation is an "Attempt," ending in either Failure (det_status = 0, -1, or -2 in the simulation DRM), or Success (det_status = 1 in the DRM).

In general, "cumulative" means the total number of detections, counting repeated detections of the same target. And "unique" means that repeat detections, or failures, are not counted.

Specific notes on row-by-row properties:

  • Attempts end in either failure or successful detection (A = F + D).
    This carries through to cumulative numbers in the table, so that:
    Attempt_cume = Fails_cume + Detected_cume
    This property does not carry through to unique counts. (E.g., two attempts on one target could end in one failure and one success, and all unique numbers would equal 1.)

  • Failures also sub-divide, so that:
    Fails_Cume = Fails_IWA + Fails_OWA + Fails_SNR
    All of the reported subcategories are cumulative counts.

  • Failure categories do not make sense for stars because there is no way to attribute a specific set of planet failures to the host star. They are blanked out in the table.

  • Detected(n) for planets means:
    "the number of planets having exactly n occurrences of (det_status=1) during the mission".
    There is also an overflow category, Det(n > 4).

  • Detected(n) for stars is less obvious. We choose it to mean:
    "the number of stars with any combination of planets having exactly n ocurrences of (det_status=1) during the mission".
    This is generally useful only for Earth-only SU's, or SU's that are restricted to one planet type.

  • Because the Detected(n) counts partition the set of all detections, we get this relationship:
    Detected_uniq = Det(n=1) + Det(n=2) + Det(n=3) + Det(n=4) + Det(n>4)

  • Note that the Detected(n) categories imply that n detections were cumulatively performed. So, provided the overflow class (Detected(n>4)) is empty, we also have:
    Detected_cume = 1*Det(n=1) + 2*Det(n=2) + 3*Det(n=3) + 4*Det(n=4)