FastJet  3.1.3
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Modules Pages
BUGS
1 Known bugs
2 ----------
3 
4 1. For N log(N) strategies requiring CGAL, in CGAL versions 3.7, 3.8,
5  and 3.9-beta1, in rare case of near degeneracies, the program may
6  abort with the following error message:
7 
8  terminate called after throwing an instance of
9  'CGAL::Assertion_exception'
10  what(): CGAL ERROR: assertion violation!
11  Expr: false
12  File: .../include/CGAL/Triangulation_2.h
13 
14  This is due to an excessively conservative check in versions 3.7,
15  3.8, and 3.9-beta1 of CGAL that has been fixed in CGAL-3.9. Note
16  also that versions<=3.6.1 do not show that behaviour. If you
17  encounter this problem, please switch to an appropriate version of
18  CGAL. [Thanks to the CGAL development team and to Olivier Devillers
19  in particular for the quick fix.]
20 
21 2. With JetMedianBackgroundEstimator, the computation of sigma()
22  implicitly assumes that all jets have the same area. The fact that
23  they don't introduces a bias at the level of a couple of tens of
24  percent. GridMedianBackgroundEstimator should not suffer from this
25  problem. Thanks to Christopher Young for bringing this to our
26  attention.
27 
28 ----------------------------------------------------------------------
29 Other issues to be aware of
30 ---------------------------
31 
32 1. Some algorithms can have ambiguities in the clustering steps and we
33  do not guarantee that different internal clustering strategies (or
34  different fastjet versions) will always resolve the ambiguities
35  identically. This issue doesn't arise with normal particle inputs.
36 
37  However:
38 
39  - Some (older) versions of Pythia 6 sometimes produce massive
40  particles with pt=0; phi is then ill-defined, while the rapidity
41  is finite and we make no guarantees about how we treat the
42  particle.
43 
44  - inputs that lie on a perfect grid in rapidity-phi (e.g. one
45  interpretation of a calorimeter) cause many interparticle
46  distances to be identical. In the kt, Cambridge/Aachen and anti-kt
47  algorithms, many dij can therefore also be identical. The choice
48  of which dij to take is then ambiguous and what fastjet actually
49  does may depend on: the compiler, the machine architecture
50  (because of rounding errors), the fastjet version and the
51  clustering strategy.
52 
53  Physically this issue should not change the jets
54  much. Nevertheless, one might choose to break any degeneracy
55  explicity, e.g. using information from the location of energy
56  deposits in each calorimeter tower, so that the inputs are not on
57  a perfect grid.
58 
59 2. For some of the plugins (listed below), the result of the clustering
60  depends on how "sort" treats equal entities. Since this is
61  supposedly undefined, FastJet has no control over the different
62  results that one may obtain using those plugins.
63 
64  - The ATLAS-Cone plugin: At the beginning of the stable-cone search,
65  the input particles are sorted in Et. In that sort, particles with
66  an Et difference below 0.001 are considered as having the same Et
67  (see line 80 of Jet.hh), resulting in the undefined behaviour
68  mentioned above.
69 
70  A consequence of this is that, if 2 input particles have too
71  similar an Et, the stable cone search will consider them in an
72  undefined order. If the 2 resulting stable cones are too close
73  (deta<0.05, dphi<0.05) one will be accepted and the other
74  rejected. Which one depends on the ordering and is thus
75  undefined. If the 2 stable cones do not have the same constituents
76  this could affect the result of the clustering.
77 
78  - In the TrackJet plugin, input particles are sorted in Et. If two
79  particles have the same Et (within machine precision), the order
80  in which they will be considered is undefined. Note however that to
81  have exactly the same pt, the particles are very likely to be
82  back-to-back. In that case, only the order of the clustering will
83  be affected but not the final jets.
84 
85  Relative to the original code, we have replaced the use of 'sort'
86  with 'stable_sort' so that in case of such a degeneracy the order
87  will be the same as that of the original particles and no
88  random behaviour is to be expected. Note however that the issue
89  could still arise if one changes the ordering of the input
90  particles.
91 
92 3. Copy of ClusterSequenceArea (and the other CS area-related classes)
93  is not fully implemented.
94 
95 ----------------------------------------------------------------------
96 Probably solved
97 ---------------
98 
99 1. Compilation issues have been reported under windows with the VC7.1
100  compiler. We have incorporated fixes proposed by Ivan Belyaev for
101  this, but do not have access to an appropriate machine to test it
102  ourselves. (NB: it in any case works with cygwin).
103 
104 
105