Youetal2025

Version of the reltrans code used in You et al (2025), Nature Communications. This version includes the model flavour reltransCpF, which outputs a lag vs frequency spectrum between two given energy bands. The code was branched from older versions of reltrans, so is kept in a separate repository for version control reasons. The genreal reltrans documentation can be found here: https://reltransdocs.readthedocs.io/en/latest/

To use reltransCpF, load reltrans into xspec like usual (as is described in the link above), then the new model flavour can be called with the "model" command within xspec. The parameters are as follows:

	Parameter	Unit	Default Value
1	h	Rg/Rh	6.0
2	a		0.998
3	inc	deg	30.0
4	rin	Rg/ISCO	-1.0
5	rout	Rg	1000.0
6	z		0.0
7	Gamma		2.0
8	logxi		3.0
9	Afe	solar	1.0
10	logNe		15.0
11	kTe	keV	300.0
12	nH	10^22	0.0
13	boost		1.0
14	Mass	Msun	10.0
15	Ea1	keV	0.3
16	Ea2	keV	1.0
17	Eb1	keV	1.0
18	Eb2	keV	4.0
19	phiAB	rad	-0.2
20	AB_exp		1.0
21	g		0.1
22	g_exp		1.0
23	norm		1.0

Parameters 1-20 are the same as in other model flavours. Parameters 21-24 define the two energy bands that the lag spectrum is taken between. The model cross spectrum is taken between band A and Band B, such that a positive lag corresponds to band A lagging band B. Band A is energy Ea1 to Ea2, and band B is energy Eb1 to Eb2. Parameters 19-22 describe the prescription for the continuum lags. The usual reltrans model parameters g and phiAB are defined for each Fourier frequency range considered (see the readthedocs). Here, they depend on frequency $\nu$ as
$\texttt{phiAB}(\nu) = \texttt{phiAB} \times \nu^\texttt{ABexp}$
$\texttt{g}(\nu) = \texttt{g} \times \nu^\texttt{gexp}.$
Finally, parameter 23 must be fixed at unity.

The epoch 1 lag vs frequency data modelled in the paper can be found in the directory HXMTdata/. The two xcm files can be used to reproduce the fits shown in Fig 6 of the paper.