nway
nway copied to clipboard
p_any cutoff
Q. The value of the p_any cutoff, calculated in the following manner, can be used again to improve the best match or just to report in a paper to show the percentage of false detections.
You can calibrate a p_any cut-off with the following steps:
- Create a offset catalogue to simulate random sky positions: nway-create-fake-catalogue.py --radius 40 COSMOS_XMM.fits COSMOS_XMM-fake.fits
- Match the offset catalogue in the same way as this run: ../nway.py COSMOS_XMM-fake.fits :pos_err COSMOS_OPTICAL.fits 0.1 --out=examplenway.fits-fake.fits --radius 20 --prior-completeness 0.9
- determining the p_any cutoff that corresponds to a false-detection rate nway-calibrate-cutoff.py examplenway.fits examplenway.fits-fake.fits
That is true; I guess it should be stated clearer in the manual?