For this analysis, completeness and reliability are defined strictly based on the
number of confirmations.
All sources with Nc > 5 are classified as 'true' sources.
Sources with Nc < 6 are classified as 'false' sources.
Nc represents the number of confirmations.
Completeness only involves sources with > 5 confirmations or 'true' sources.
C = 1 -- (number_missed_scans/total_scans_available)
Reliability refers to all sources.
R = number_true/(number_true + number_false/10)
number_true = all sources with Nc > 5
number_false = number of extractions of sources with Nc < 6
Figure 1: Plot of completeness and reliability vs. K magnitude
for the June 1994 scans of the M92 area.
Problem Summary for Incomplete and Unreliable Sources
M92 June 1994
(K magnitude 
)
---------------------------------------------------------------------
897 sources output from confirmation processor; 8970 sightings
20 unreliable sources (Nc < 6)Due to sky:
8 are galaxies
1 is radiation hit
Due to easily fixable problems:
5 are persistence (just outside 2" cleanup window - R1 problem)
3 are bright source (k < 8) position problems > 3" separation
(3.8", 3.3" and 3.0003")
3 are R1 merge problems
12 incomplete sources (5 < Nc < 10)
Due to sky:
6 are galaxies
3 are double star detections 2.6" 2.3" and 4.6"
Due to easily fixable problems:
3 are bright source (k < 8) position problems > 3" separation (3.8", 3.3" and 3.0003")
total: 32 unreliable or incomplete sources
involving 16 real sources
