Five-day MAXI and BAT data show opposite soft–hard X-ray correlations in 4U 1636−536 in 2014–2020 and 2020–2026

We test the relation between low- and high-energy X-rays from 4U 1636−536 with public daily MAXI and Swift/BAT data. We form five-day means from common dates. Two equal 2,190-day intervals contain 354 paired bins from 2014-03-23 to 2020-03-21 and 286 from 2020-03-21 to 2026-03-20; end dates are excluded. The Pearson correlation of MAXI 2–4 keV with BAT 15–50 keV changes from −0.300 to +0.439. Approximate 95% intervals from 100-day block sampling are [−0.382, −0.215] and [+0.220, +0.602]. All three early 730-day blocks have negative point estimates; all three late blocks have positive point estimates. The signs remain under the tested changes in bin width and daily weights. The early negative result also remains after removal of older MAXI camera data. Slow changes account for much of the later positive relation. After removal of a local mean within 100 days on each side, the later correlation is +0.131, with an interval of [−0.075, +0.314]. This test does not establish a positive relation in the residual variation. The interval split was chosen after data inspection. Unequal coverage, measurement noise, and calibration history limit interpretation. No new period, exact change date, or physical cause is established. We preserve four monitor snapshots, source hashes, paired data, all processing checks, figures, code, and reproduction instructions.

4U 1636-536 · neutron star · X-ray binary · MAXI · Swift BAT · long-term variability · correlation · public data · reproducible analysis