Report for Simulation #4

The bottom line is that we get 99% completeness, with most of the missing QSOs in the 2.5 - 3.5 redshift range. In this range, the completeness drops to 92%. Correcting for the fact that ten times as many QSOs were put into this simulation than are expected from the star counts, the efficiency is 55%. Of the selected candidates, 55% are QSOs, 19% are NELGs, 3% are white dwarfs, and 23% are galactic stars. This assumes that the simulations are a fair representation of the sky.

Okay, I have successfully run the fourth simulation through the QSO target selection code. The procedures which convert Xiaohui's format into survey inputs and also which print out the figure of merit had to be slightly altered to handle the larger data set. In the future, the input data set should be broken up into fields (of maybe 1000 objects each). In the input ASCII file, a new field is signaled by inserting three lines: (1) a blank line, (2) a line containing only: "NEW FIELD", and then (3) another blank line. Another small change that I would like to request in the simulations is that the keymag fields should contain the magnitudes without reddening and without errors. They currently have the same numbers as the psf magnitude fields.

This is the figure of merit:

                                        # targets/#objects brighter than 20.015 # objects

Targets selected (by type):     1       59/1510 3490
                                2       68/9958 10043
                                3       11/2332 2332
                                4       0/0     0
                                5       0/0     0
                                6       0/0     0
                                7       0/0     0
                                8       113/114 384
                                9       17/22   58
                                10      3342/3373       8206
                                        -----------------
                                        3610/17309      24513

QSO completeness (z):           -10 - 1 1212/1212       1813
                                1 - 1.5 787/789 2255
                                1.5 - 2 600/600 1802
                                2 - 2.5 402/403 1231
                                2.5 - 3 221/236 673
                                3 - 3.5 75/86   292
                                3.5 - 4 33/35   99
                                4 - 6   12/12   41
                                6 - 10  0/0     0
                                        -----------------
                                        3342/3373       8206

QSO completeness (m):           -10 - 15        4/4     4
                                15 - 16 3/3     3
                                16 - 17 29/29   29
                                17 - 18 185/187 187
                                18 - 19 1046/1051       1051
                                19 - 20 2038/2060       2060
                                20 - 30 37/39   4872
                                        -----------------
                                        3342/3373       8206

Because this simulation included objects fainter than the limiting magnitude (i = 20), I added an extra column. The completeness I quote will be the completeness including only those objects up to the limiting magnitude. The new (last) column in the figure of merit is the total number of objects without including those below the threshold. The reason there are some detected objects in the magnitude 20 - 30 category is that there is 0.015 magnitudes of reddening, so since the keymag doesn't take this out (and this is used to bin the data), some objects close to threshold show up just over i=20.