What the PHOX codes cannot do  General
As any computer code, the PHOX codes obey Feynman's
`GIGO' principle:
"Garbage In, Garbage Out":
if you ask the wrong question to a PHOX code, you may get a
wrong answer.
Here we list the two most important warnings regarding the use of the
PHOX codes:
 The PHOX codes
ARE NOT FULL EVENT GENERATORS.
The terminology "Monte Carlo program" refers here to the numerical
integration technique used. The user should be careful not to be confused
by this terminology.
 The PHOX codes
do not provide a full, exclusive portrait of events
which could for example be further processed through a detector
simulation. This is not their purpose.
 The PHOX codes are not
designed to be interfaced with parton showers
and hadronisation models.
The PHOX codes are instead
self contained
programs designed to
compute sufficiently inclusive observables
with the possibility to account for
various selection cuts
which can be formulated at the parton level.
The virtue of the PHOX codes is to provide predictions of
Next to Leading Order (NLO)
accuracy in the pertubative expansion in powers
of alpha_{s} for such inclusive observables.
This means that the PHOX codes give more accurate and
quantitative answers than a full event generator regarding such inclusive
quantites, but at the expense of a more restricted range of applicability.
The PHOX codes and full event generators are
complementary to each other.
 The PHOX codes SHARE ALL THE
LIMITATIONS INHERENT IN ANY FIXED ORDER PERTURBATIVE CALCULATION
based on the QCDimproved parton model.
In particular, the results provided by the PHOX codes should be
considered with circumspection if
 the relevant partonic energy scale
involved in the subprocess
is not large enough
compared to the typical hadronic scale O(1 GeV)
(these codes involve no primordial k _{ T } smearing nor any
highertwist/power correction).

the relevant partonic energy scale
is much smaller than the
C.M.S. energy of the collision
(these codes do not account for smallx physics).

the observable considered is infrared sensitive
(in the calculation of a distribution at any fixed order in perturbation
theory, this infrared sensitivity is indicated by the appearence of
large logarithms depending on the ratio of widely different scales.
Such large logarithms are associated with soft gluon effects. These codes
do not (yet) resum these multiple soft gluon effects. Therefore their
predictions are not reliable whenever such large terms appear  example:
transverse momentum (Q _{ T }) distribution of a photon pair with
a large invariant mass M, in the region where Q _{ T } / M << 1).
In practice, the actual use of the various PHOX codes
may not be affected by these restrictions, depending on the type of process
concerned. For more detailed information please consult the warnings
specific to the particular PHOX code which you intend to use.
Back to main page