]>
Commit | Line | Data |
---|---|---|
21886bb6 | 1 | * |
2 | * $Id$ | |
3 | * | |
4 | * $Log$ | |
5 | * Revision 1.1.1.2 1996/10/30 08:27:44 cernlib | |
6 | * Version 7.04 | |
7 | * | |
8 | * Revision 1.1.1.1 1996/04/12 15:29:09 plothow | |
9 | * Version 7.01 | |
10 | * | |
11 | * | |
12 | #include "pdf/pilot.h" | |
13 | C Version 2 CTEQ distribution function in a parametrized form. | |
14 | C By: J. Botts, H.L. Lai, J.G. Morfin, J.F. Owens, J. Qiu, W.K. Tung & H. Weerts | |
15 | C | |
16 | C To avoid the proliferation of parton distribution functions, we recommend | |
17 | C that | |
18 | C these distributions should replace Version 1 CTEQ distributions for all | |
19 | C general | |
20 | C usage. The differences between the two sets of distributions, as briefly | |
21 | C described below, do not significantly affect most applications for | |
22 | C fixed-target | |
23 | C and hadron-collider applications. | |
24 | C | |
25 | C Both CTEQ1 and CTEQ2 distributions fit existing DIS, Drell-Yan and Direct | |
26 | C photon | |
27 | C data with excellent chi-squares. They represent two distinct ways to resolve | |
28 | C the inconsistency between CCFR and NMC measurements on F2 at small-x and the | |
29 | C neutrino dimuon measurements of s(x), as discovered by the CTEQ1 analysis: | |
30 | C in the CTEQ1 analysis, the F2 measurements of CCFR and NMC are taken | |
31 | C seriously, | |
32 | C leaving out the dimuon information on s(x); whereas in the CTEQ2 analysis, | |
33 | C the | |
34 | C latest CCFR NLO dimuon analysis of s(x) is used (within errors) as input | |
35 | C while | |
36 | C leaving out the small-x F2 data which conflict with this input. The small-x | |
37 | C behavior of the parton distributions are contrained in the latter case by the | |
38 | C newly released HERA data. For details, see our forthcoming paper. | |
39 | C | |
40 | C This file contains three versions of the same CTEQ2 parton distributions: | |
41 | C | |
42 | C Two "front-end" subprograms: | |
43 | C FUNCTION Ctq2OPf (Iset, Iparton, X, Q, Irt) | |
44 | C returns the PROBABILITY density for a GIVEN flavor; | |
45 | C SUBROUTINE Ctq2OPs (Iset, Pdf, XX, QQ, Irt) | |
46 | C returns an array of MOMENTUM densities for ALL flavors; | |
47 | C One lower-level subprogram: | |
48 | C FUNCTION Ctq2OPd (Iset, Iprtn, XX, QQ, Irt) | |
49 | C returns the MOMENTUM density of a GIVEN valence or sea distribution. | |
50 | C Supplementary functions to return the relevant QCD parameters and other | |
51 | C information concerning these distributions are also included (see below). | |
52 | C | |
53 | C Since this is an initial distribution of version 2, it is | |
54 | C useful for the authors to maintain a record of the distribution list in | |
55 | C case there are revisions or corrections. | |
56 | C In the interest of maintaining the integrity of this package, | |
57 | C please do not freely distribute this program package; instead, refer any | |
58 | C interested colleagues to direct their request for a copy to: | |
59 | C Botts@hades.ifh.de or Lai@cteq11.pa.msu.edu | |
60 | ||
61 | C If you have detailed questions concerning these CTEQ2 distributions, direct | |
62 | C inquires to Botts, Lai (see above) or Wu-Ki Tung (Tung@msupa.pa.msu.edu). | |
63 | ||
64 | C ------------------------------------------- | |
65 | C Detailed instructions follow. | |
66 | ||
67 | C Name convention for CTEQ distributions: CTEQnSx where | |
68 | C n : version number (currently n = 2) | |
69 | C S : factorization scheme label: = [M D L] for [MS-bar DIS LO] | |
70 | C resp. | |
71 | C x : special characteristics, if any | |
72 | C (e.g. S(F) for singular (flat) small-x, L for "LEP lambda value") | |
73 | ||
74 | C Explanation of functional arguments: | |
75 | ||
76 | C Iset is the set label; in this version, Iset = 1, 2, 3, 4, 5, 6 | |
77 | C correspond to the following CTEQ global fits: | |
78 | ||
79 | C cteq2M : best fit in the MS-bar scheme | |
80 | C cteq2MS : singular small-x | |
81 | C cteq2MF : flat small-x | |
82 | C cteq2ML : large lambda (Lambda(5) = 220 MeV) | |
83 | ||
84 | C cteq2L : best fit in Leading order QCD | |
85 | ||
86 | C cteq2D : best fit in the DIS scheme | |
87 | ||
88 | C Iprtn is the parton label (6, 5, 4, 3, 2, 1, 0, -1, ......, -6) | |
89 | C for (t, b, c, s, d, u, g, u_bar, ..., t_bar) | |
90 | ||
91 | C X, Q are the usual x, Q; | |
92 | C Irt is a return error code (see individual modules for explanation). | |
93 | C | |
94 | C --------------------------------------------- | |
95 | C --> QCD parameters for parton distribution set Iset can be obtained inside | |
96 | C the user's program by: | |
97 | C Dum = PrCtq2O | |
98 | C > (Iset, Iord, Ischeme, MxFlv, | |
99 | C > Alam4, Alam5, Alam6, Amas4, Amas5, Amas6, | |
100 | C > Xmin, Qini, Qmax, ExpNor) | |
101 | C where all but the first argument are output parameters. | |
102 | C They should be self-explanatory -- see details under ENTRY PrCtq2O. | |
103 | ||
104 | C Since the QCD Lambda value for the various sets are needed more often than | |
105 | C the other parameters in most applications, a special function | |
106 | C Wlamd2O (Iset, Iorder, Neff) is provided | |
107 | C which returns the lambda value for Neff = 4,5,6 effective flavors as well as | |
108 | C the order these values pertain to. | |
109 | ||
110 | C ---------------------------------------------- | |
111 | C The range of (x, Q) used in this round of global analysis is, approxi- | |
112 | C mately, 0.01 < x < 0.75 ; and 4 GeV^2 < Q^2 < 400 GeV^2 for fixed target | |
113 | C experiments and 0.0001 < x < 0.01 from first official data of HERA. | |
114 | ||
115 | C The range of (x, Q) used in the reparametrization of the QCD evolved | |
116 | C parton distributions is 10E-5 < x < 1 ; 1.6 GeV < Q < 1 TeV. The | |
117 | C functional form of this parametrization is: | |
118 | ||
119 | C A0 * x^A1 * (1-x)^A2 * (1 + A3 * x^A4) * [log(1+1/x)]^A5 | |
120 | ||
121 | C with the A'coefficients being smooth functions of Q. For heavy quarks, | |
122 | C a threshold factor is applied to A0 which simulates the proper Q-dependence | |
123 | C of the QCD evolution in that region according to the renormalization | |
124 | C scheme defined in Collins-Tung, Nucl. Phys. B278, 934 (1986). | |
125 | ||
126 | C Since this function is positive definite and smooth, it provides sensible | |
127 | C extrapolations of the parton distributions if they are called beyond | |
128 | C the original range in an application. There is no artificial boundaries | |
129 | C or sharp cutoff's. | |
130 | C ------------------------------------------------ | |
131 | ||
132 | FUNCTION Ctq2OPf (Iset, Iparton, X, Q, Irt) | |
133 | ||
134 | C This function returns the CTEQ parton distributions f^Iset_Iprtn/proton | |
135 | C --- the PROBABILITY density | |
136 | ||
137 | C (Iset, Iparton, X, Q): explained above; | |
138 | ||
139 | C Irt : return error code: see module Ctq2OPd for explanation. | |
140 | ||
141 | C IMPLICIT DOUBLE PRECISION (A-H, O-Z) | |
142 | C+SEQ, IMPDP. | |
143 | ||
144 | Ifl = Iparton | |
145 | JFL = ABS(Ifl) | |
146 | C Valence | |
147 | IF (Ifl.Eq.1 .or. Ifl.Eq.2) THEN | |
148 | VL = Ctq2OPd (Iset, Ifl, X, Q, Irt) | |
149 | ELSE | |
150 | VL = 0. | |
151 | ENDIF | |
152 | C Sea | |
153 | SEA = Ctq2OPd (Iset, -JFL, X, Q, Irt) | |
154 | C Full (probability) Distribution | |
155 | Ctq2OPf = (VL + SEA) / X | |
156 | ||
157 | Return | |
158 | C ************************* | |
159 | END |