A day reflection test did not detect a short gap time direction in two FAST burst catalogs

We tested the order of short arrival time gaps in two public FAST fast radio burst catalogs. The catalogs contain 1,652 bursts from FRB 20121102A and 1,863 bursts from FRB 20201124A. For successive triples, we used the score (b-a)/(b+a), where a and b are the first and second gaps. This score changes sign under time reversal. Reflection preserves all pair distances within a day. We summed scores within integer catalog MJD days and used random signs on whole day score vectors. The test assumes independent days and a symmetric law under day reflection. We fixed gap cutoffs at 0.1, 0.3, and 1.0 seconds and adjusted for their joint search. Only 60 and 18 successive triples have both gaps at most one second. The source p-values are 0.991435 and 0.907170. Both become 1.0 after Holm adjustment across the two sources. Of 455 short groups, 387 contain exactly two bursts and have no internal unmarked gap order. Checks with other blocks and finer Xu waiting times give the same test decision. A conditional simulation shows that the small number of longer groups can hide modest orientation preferences. This result gives no detection of a net short gap direction in the tested score. It does not establish random emission or exclude burst triggering. The package contains input data, source records, code, event ledgers, figures, and a full offline reproduction check.

fast radio bursts · FAST · arrival times · time reversal · gap order · block randomization · reproducible research