-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCHANGELOG
158 lines (89 loc) · 4.47 KB
/
CHANGELOG
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
Changelog:
--
Modifications needed for Python3.
--
Apply slips to synthetic data, not observational data,
so that we see always the same o. d.
--
Jet has higher priority than nebula.
Envelope has lower priority than nebula.
exit(1) in subroutine roche() if Newton-Raphson slipped
Slips are initially set to zero in zero_slips().
Barycentric velocity correction is applied to IF data
(see velocities.html in Astropy and Wright & Eastman 2014).
--
Shellspec50 and corresponding changes.
Radial velocity field (a.k.a. disk wind) was added.
Asymmetric jet (w. asymjt parameter) was added.
--
template.in was updated ("dd [pc]" must not contain a space)
Fitting of SED and SPE data did not work, because sedsyn['wavesyn']
and sedsyn['fluxsyn'] arrays were not properly initialized.
--
Normalized spectra ("SPE") are also computed and chi^2 evaluated
(see e.g. Spectra.dat). It is actually a minor modification of
SED; the corresponding class SPEData is inherited from SEDData.
Hermite polynomials are used for interpolation of spectra.
An important correction in Pyterpol, namely PHOENIX absolute
synthetic spectra, which were computed for vacuum wavelength,
but we need air wavelengths in Pyshellspec!
--
Spectral-energy distribution (SED) is now computed, Shellspec is
called in parallel for different phases; corresponding observations
can be also input (see e.g. Sed.dat) and chi^2 evalued.
Semiamplitude K [km/s] was added to nebula and disk objects.
--
Differential visibility dif |V| is now computed (if use_differential=True).
Visibility amplitude |V| and phase arg V observables were added both
to observations.py and interface.py.
Shadowing was implemented in Shellspec; optically-thin scattering thus
can be modelled. Corresponding free parameters can be fitted.
--
Linear limb-darkening coefficients are interpolated for Teff, logg, Z
and all wavelengths (Van Hamme 1993).
Nebula temperature inversion (tinvnb) is now linear from hinvnb up to aneb;
no more jumps which may create artefacts in low-resolution models.
--
dry_run and overwrite keywords were added to interface,
to make debugging and restart of hi-res simulations feasible.
tempif* directories are created for each wavelength even if ncpu = 1.
--
Weigthing of |V^2| data is 1 (for VEGA), if no |T_3| are present.
On the other hand, individual contributions to chi^2 are always
multiplied by 0.5 factor, if both |V^2| and |T_3| are present.
--
asini = sma*dinc parameter is now fixed instead of sma
Ufo (i.e. low-priority disc) object was added to model extended
disc atmosphere, or scattered light (when ithom = iray = 0).
Flow object was added to model off-center jet structures.
--
New shellspec43.f was used (with yet better optically thick transfer).
Nevertheless, one has to be still aware of artefacts arising from
sharp boundaries between thin/thick objects (like slab+ufo);
see large contrib. to contribution function cfi in split2() subroutine.
An exit from a thread now exits the main thread too.
hmulnb parameter was added (scale-height multiplication factor for nebula).
--
Another increase of modp = 10000000.
--
Individual chi^2 terms are output in *.syn.dat files.
New shellspec39.f was used (with modified optically thick transfer).
For |V^2| and |T_3|, the chi^2 is computed as 1/2 chi^2_V2 + 1/2 chi^2_T3.
Nebula object was added (i.e. flared disk with exponential vertical profile).
UV filters of OAO2 were added.
Comments in observations ASCII files are allowed (removal of wrong data).
Pyterpol synthetic spectra are used instead of black-body approximation.
Line transfer is used in particular for the H_alpha wavelength.
fit.log is of higher precision and contains numbers of observations.
temp* directories are for ALL effective wavelengths (not for cpus);
otherwise it is not possible to debug the multi-cpu code!
Old lightcurve, shellspectrum, shellspec.out files are properly deleted,
and pyterpol, starspec1, starspec2, line.dat files are symlinked.
modp=1000000 is used in shellspec39.f to deal with optically thick medium.
|T_3| for some datasets can be excluded (None values) independently of arg T_3.
Output lightcurve (*.syn.dat) contains effective bandpass and offset.
Bandpass fluxes [in J s^-1 m^-2] and UBVR magnitudes are correctly calculated,
so offsets are not needed anymore and SED can be fitted.
Exitcodes in shellspec were added, as well as correspoding changes in pyshellspec.
Checking of the wavelength range of synthetic spectra in shellspec.
--