<pre> ./<a href="#robo_top_of_doc">results.pl</a> or from <a href="./launch.html#robo_top_of_doc">launch</a> or <a href="./finishnumerics_pl.html#robo_top_of_doc">finishnumerics.pl</a>
</pre>
<p class="item_name">USES</p>
<pre> $paramfile, <a href="./perlsrc/header_pm.html#robo_top_of_doc">header.pm</a>, <a href="./perlsrc/getinfo_pm.html#robo_top_of_doc">getinfo.pm</a>, <a href="./perlsrc/dirform_pm.html#robo_top_of_doc">dirform.pm</a> *x*.out in the leaf directories,
-<p>Generated from ./general/launchcleangraph with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/launchcleangraph with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> integrates all integrands for the specified order of epsilon
as one function, instead of individually integrating different
pole structures and numbers of integration variables separately.
this is particularly useful when intermediate calculations produce
large positive and negative outputs, which cancel to give a small
final result. NB if specifying a negative order, this must be expressed
as 'm1', instead of '-1'.
OPTIONS
to use a param.input file with a different name
use option "-p paramfile"
to specify a different directory to work in
use option "-d workingdirectory"
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/dotogether.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/dotogether.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> launches the calculation. Calculation runs until it hits an error, or completes to the
level specified with the 'exe' flag in paramfile
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/cleanall.sh with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/cleanall.sh with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> To create optimized fortran code for any functions left implicit
at the algebraic level, to be used during the numerical integration
</pre>
<p class="item_name">INPUTS</p>
<pre> $paramfile (default is param.input), read via module header
RESULT
-copies the file 'dummytofortran.m', and edits it to read the
files for functions specified in the parameter file
-launches dummytofortran.m, which creates optimized fortran code
for these functions and puts them in the desired directory.
OPTIONS
to use a param.input file with a different name
use option "-p paramfile"
to use a template file with the a different name
use option "-t templatefile"
to specify a different directory to work in
use option "-d workingdirectory"
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/createdummyfortran.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/createdummyfortran.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> the file '<a href="#robo_top_of_doc">batch</a>' acts as a template for <a href="../subexp_pl.html#robo_top_of_doc">subexp.pl</a> to create files
subdir/graph/<a href="#robo_top_of_doc">batch</a>[polestructure], which are run/submitted to perform
subtraction & eps expansion via subandexpand[polestructure].m, and
the creation, compilation and execution of all files needed for numerical
integration via <a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>. From <a href="../justnumerics_pl.html#robo_top_of_doc">justnumerics.pl</a>, <a href="#robo_top_of_doc">batch</a>[polestructure]
only does the creation, compilation and execution of all files needed for
numerical integration via <a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>.
-<p>Generated from ./general/perlsrc/batch with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/batch with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> parameters parsed from <a href="../launchcleangraph.html#robo_top_of_doc">launchcleangraph</a> via ARGV:
dirbase: the polestructure subdirectory to be cleaned
allflag: specifies whether all files are to be removed
fullflag: when a full clean is specified, this flag states whether
it is the polestructure subdirectories, or the graph directory, to be cleaned
RESULT
intermediate files removed
</pre>
<p class="item_name">SEE ALSO</p>
<pre> <a href="../launchcleangraph.html#robo_top_of_doc">launchcleangraph</a> (created by <a href="../subexp_pl.html#robo_top_of_doc">subexp.pl</a>, <a href="../justnumerics_pl.html#robo_top_of_doc">justnumerics.pl</a>)
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/perlsrc/cleanup.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/cleanup.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makeredo.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makeredo.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makesub.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makesub.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/header.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/header.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makeint.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makeint.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> is called by <a href="../finishnumerics_pl.html#robo_top_of_doc">finishnumerics.pl</a>, <a href="../justnumerics_pl.html#robo_top_of_doc">justnumerics.pl</a>, <a href="../results_pl.html#robo_top_of_doc">results.pl</a>, <a href="../subexp_pl.html#robo_top_of_doc">subexp.pl</a> via getinfo::routinename(filename)
</pre>
<p class="item_name">USES</p>
<pre> various filenames parsed to its subroutines
</pre>
<p class="item_name">PURPOSE</p>
<pre> collects the subroutines which are used to read various pieces of information from output/intermediate files
</pre>
<p class="item_name">INPUTS</p>
<pre> arguments:
infofile/logfile/graphfile/resfile: name of file where required information is held
RESULT
the required piece of information is return from the subroutine
-<p>Generated from ./general/perlsrc/getinfo.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/getinfo.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makemakerun.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makemakerun.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> contains subroutines which are used to format strings into valid
directory syntax
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/perlsrc/dirform.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/dirform.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makesum.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makesum.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/preparenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/preparenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/polenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/polenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/finishpolenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/finishpolenumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> is called by <a href="../subexp_pl.html#robo_top_of_doc">subexp.pl</a>, <a href="../justnumerics_pl.html#robo_top_of_doc">justnumerics.pl</a>, <a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>, <a href="../finishnumerics_pl.html#robo_top_of_doc">finishnumerics.pl</a>, *subfile.pl
-<p>Generated from ./general/perlsrc/launchjob.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/launchjob.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makejob.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makejob.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/makemake.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/makemake.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/perlsrc/writefiles.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/writefiles.pm with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> is called from <a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>, subexploop.pl, subexpgeneral.pl, finishnumericsloop.pl and finishnumericsgeneral.pl,
justnumericsloop.pl and justnumericsgeneral.pl
USED BY
<a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>, subexploop.pl, subexpgeneral.pl, finishnumericsloop.pl and finishnumericsgeneral.pl,
justnumericsloop.pl and justnumericsgeneral.pl
</pre>
<p class="item_name">PURPOSE</p>
<pre> compiles libbases.a or libcuba.a in a directory where it can then be used to perform the numerical integration.
separate directories can be needed for each pole structure, since the compilation of the executables
may be performed on different machines for different pole structures. This is also true for different calculations
running concurrently. When the input guarantees the program will be run fully locally, the library is made only once, and
shared by the different pole structures.
</pre>
<p class="item_name">INPUTS</p>
<pre> parameters parsed via ARGV:
integpath: path where all the source files for numerical integrator can be found
graph: name of graph
polestruct: either the pole structure, or a flag stating that the program is running locally
compiler: compiler to compile library with
makefile: makefile used to compile library
RESULT
numerical integrator is compiled in a directory ready to be used for numerical integration
</pre>
<p class="item_name">SEE ALSO</p>
<pre> <a href="./polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a>, subexploop.pl, subexpgeneral.pl, finishnumericsloop.pl and finishnumericsgeneral.pl,
justnumericsloop.pl and justnumericsgeneral.pl
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/perlsrc/remakebases.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/remakebases.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> finds the path for Mathematica, and then runs the inputfile.m,
with any output directed to outputfile
</pre>
<p class="item_name">INPUTS</p>
<pre> parameters parsed via ARGV:
@ARGV=(inputfile.m, outputfile)
RESULT
inputfile.m is run, output directed to outputfile
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/perlsrc/mathlaunch.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/perlsrc/mathlaunch.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> doing the iterated sector decomposition in Mathematica
</pre>
<p class="item_name">INPUTS</p>
<pre> $paramfile (default is param.input), read via module header
$templatefile (default is Template.m)
RESULT
-adds decomposition part to the file $currentdir/$graph.m
-launches the decomposition and writes the results to
$currentdir in the form of lists containing the
parametric functions for each pole structure
OPTIONS
to use a param.input file with a different name
use option "-p paramfile"
to use a template file with the a different name
use option "-t templatefile"
to specify a different directory to work in
use option "-d workingdirectory"
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/decompose.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/decompose.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> completes the calculation from where it was left off (specified via $exe in paramfile)
</pre>
<p class="item_name">INPUTS</p>
<pre> $paramfile (default is param.input) read via module header
RESULT
rest of the calculation is run/submitted to be run, numerical result is then displayed
where possible, or available via 'perl <a href="./results_pl.html#robo_top_of_doc">results.pl</a> -d workingdirectory -p paramfile' once all integrations
-<p>Generated from ./general/finishnumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/finishnumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/subexp/subandexpand.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/subandexpand.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> Takes an optimized expression from <a href="./formfortran_m.html#robo_top_of_doc">formfortran.m</a>, and writes it in fortran syntax to the appropriate subdirectory.
</pre>
<p class="item_name">INPUTS</p>
<pre> from formfortran.m:
inlist: the optimized expression to be turned into fortran syntax
varletter: the letter each intermediate expression in the optimization will be prefixed by
funcname: the name of the function to be written
outfile: where the file is to be written
fortranstring1,2,3: strings with the necessary fortran syntax
xstring: string with the necessary assignments for x (IBP only), in fortran format.
RESULT
the f*.f file is written in fortran syntax to the specified subdirectory
-<p>Generated from ./general/src/subexp/ExpOpt.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/ExpOpt.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/subexp/formfortran.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/formfortran.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/subexp/symbsub.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/symbsub.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/subexp/parts.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/parts.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/subexp/formindlist.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/subexp/formindlist.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/deco/decomposition.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/decomposition.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/deco/prepareinput.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/prepareinput.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
countvars[cvf,cfv]: finds the maximum power of a set of variables cfv
in a function cvf.
findpossset[fexpr]: finds all minimal sets (wrt length) of the
integration variables which nullify a function fexpr.
returns these sets, together with an integer which
is the maximum power of these integration variables in the function.
howmanynullify[posset,integrands]: counts the number of integrands
a given set posset nullifies.
findbestset[integrands,n]: finds the 'best' set which nullifies
function n from the list of integrands.
'best' is the set which nullifies the most integrands,
and has the lowest maximum power of integration variables in
integrand n
paralleldecompose[allsecs]: maps 'decompose' onto each sector in allsecs,
uses parallel processing where available/appropriate. allsecs is a lists
of sectors to be decomposed. Returns a list of fully decomposed sectors.
exdecompose[esec,set,varnum]: performs the explicit replacements needed
for decomposition. Factorizes the integrands and recalculates the
exponents of the integration variables. Returns a sector which may
require further decomposition.
factorizeintegrand[{integ,expo,flag},tt,var]: pulls out any factors of
tt[var] in int, recalculates the exponent of tt[var] and returns
{factorized int, expo, new flag}. flag/new flag = A or B.
decompose[dsec]: Checks whether dsec needs decomposition. If not (ie all
flags = A), returns {dsec}. Otherwise the set of variables to decompose
is selected by findbestset, and exdecompose is mapped onto this set.
newsecs is the list of transformed sectors, and then this list of sectors
is further decomposed by paralleldecompose[newsecs]. when this process
terminates, decompose returns a list of fully decomposed sectors.
for preparing and writing output:
polecheck[expos]: classifies a a list of exponents by number of logarithmic,
linear and higher poles.
makesectoroutput[sector]: classifies a sector by its pole structure.
increments the counter polecount[polestruct] for this classification,
and adds a pointer to sector in a list seclist[polestruct] linked to
this classification. If polestruct is the first of this structure,
it is added to the list structlist of observed pole structures.
writeoutput[polestruct]: Updates the values of maxpoles where applicable.
the list of pointers is used to form the final
list of sectors for this polestruct to write for output. The output is
written to the appropriate [integral]P[polestruct].out. The information
on number of sectors with this pole structure is
added to the string poleinfostring
writeinfo: writes the necessary information to file [integral]OUT.info.
this info includes number of integration variables nmax, total number of
sectors produced allcount, maximum poles maxlogpole,
maxlinpole, maxhpole, and the order of the constant prefactor prefacord.
the constant prefactor constpre is written to the file prefactor.m
makeoutput: calculates the total number of sectors, performs makesecoutput
to each sector, reformats the sectors to be in the desired output format,
sorts the structlist into order of complexity, performs writeoutput for each
pole structure, then performs writeinfo.
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/src/deco/SDroutines.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/SDroutines.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> contains routines used to simplify expressions containing square roots.
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/src/deco/sqrtsimp.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/sqrtsimp.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> extracts prefactors not dependent on integration variables.
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/src/deco/constprefac.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/constprefac.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
output of collecteps in <a href="./prepareinput_m.html#robo_top_of_doc">prepareinput.m</a>
</pre>
<p class="item_name">PURPOSE</p>
<pre> for variable x with singularities at both x->0 and x->1, the integration is split
in two at x=1/2, such that all singularities now occur at x'->0
NOTE
list of indices of variables to be split has to be given in Template.m
spi is index of variable which will be split
Nn is global
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/src/deco/split.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/deco/split.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
-<p>Generated from ./general/src/util/Format.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/src/util/Format.m with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
as a template - <a href="./perlsrc/batch.html#robo_top_of_doc">batch</a>
</pre>
<p class="item_name">PURPOSE</p>
<pre> if different phase space points are required for the same graph, <a href="#robo_top_of_doc">justnumerics.pl</a>
will use the f*.f already made using <a href="./subexp_pl.html#robo_top_of_doc">subexp.pl</a>, and follow the same route as <a href="./subexp_pl.html#robo_top_of_doc">subexp.pl</a>
from this point.
</pre>
<p class="item_name">INPUTS</p>
<pre> $paramfile (default is param.input) read via module header
RESULT
if $exe>=0, <a href="./perlsrc/batch.html#robo_top_of_doc">batch</a>*l*h* and where applicable job*l*h* are created.
if $exe==1, this is not applicable, so $exe is taken to be 4
if $exe>=2, intermediate files are created by <a href="./perlsrc/polenumerics_pl.html#robo_top_of_doc">polenumerics.pl</a> (from <a href="./perlsrc/batch.html#robo_top_of_doc">batch</a>*l*h*)
if $exe>=3, .exes are made for the numerical integrations
-<p>Generated from ./general/justnumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/justnumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
<pre> Automates the calculation of multiple numeric points for a given integrand
</pre>
<p class="item_name">INPUTS</p>
<pre> multiparameter file (default is multiparam.input)
RESULT
-produces an input file for each specific point
-runs/submits to a <a href="./perlsrc/batch.html#robo_top_of_doc">batch</a> system the numerical integration for each point
OPTIONS
to use a multiparam file with a different name
use option "-p multiparamfile"
to specify a different directory to work in
use option "-d workingdirectory"
to collect results files (only required when run on the <a href="./perlsrc/batch.html#robo_top_of_doc">batch</a> system), run with
arguement "1"
to remove all derived paramfiles, run with the arguement "2", eg
./multinumerics -d demos/ -p multi.input 2
</pre>
</div> <!-- content -->
<div id="footer">
-<p>Generated from ./general/multinumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Thu Dec 20 2012 18:28:11
+<p>Generated from ./general/multinumerics.pl with <a href="http://www.xs4all.nl/~rfsber/Robo/robodoc.html">ROBODoc</a> V4.99.41 on Tue Oct 01 2013 11:54:07
+ - parallelization of the iterated sector decomposition reintroduced
+
01.10.2013
SVN rev 251 (SecDec-2.1.3)
CUBA-3.2 added
loop+general directory:
- all files updated which are necessary for using CUBA version 3.1 and above (default is set to CUBA 3.2)
- removed automated writing of f*.m files
loop directory:
- added insertion of package Fraction to multinumerics*.pl
general directory:
- bug fixed when trying to integrate constants or functions
with just one parameter but a numerical integrator which
supports a minimum of 2 integration parameters
(formfortran.m altered as well as getinfo.pm, makesub.pm)
polenumerics.pl and preparenumerics.pl, dotogether.pl)
06.08.2013
SVN rev 250 (SecDec-2.1.2)
loop directory:
- bug fixed where the integrator Vegas was not chosen when using the together flag and integrating over just one Feynman parameter.
12.07.2013
SVN rev 249 (SecDec-2.1.2)
loop directory:
- comments in param*.input changed regarding the list of s_ij invariants for pentagons and hexagons
- changed treatment of old "together" folders when using togetherflag=1 and Cpp
19.06.2013
SVN rev 248 (SecDec-2.1.1)
loop directory:
- bug fixed in subexploop.pl and subexpuserdefined.pl which occurred when the exponent of F or U contained a fraction
17.06.2013
SVN rev 247 (SecDec-2.1.1)
loop & general directory:
1) Empty lines are now removed from multiparam.input to evade read-in problems when using multinumerics*.pl
17.06.2013
SVN rev 246
loop & general directory:
1) Inserted Fraction.pm into perlsrc to be able to automatically evaluate fractions given in the input files.
2) Kinematic invariants sij, pi^2 and mi^2 which are inserted as fractions are now correctly evaluated in Perl
3) Version print was changed to SecDec-2.1.1
14.06.2013
SVN rev 245
1) Bug in formContour(P)C.m corrected which occurred when onshell-conditions were inserted in decimal form
2) Added one file needed for the installation of CUBA + revised CUBA makefile
12.06.2013
SVN rev 244
1) Fixed a bug in resultsloop.pl and resultsuserdefined.pl which occurred when using rescale=1 and prefactorflag=0
2) Added most recent version of SecDec-2.0 paper to documentary
06.06.2013
SVN rev 242
1) Improved default parameters for Suave in default param*.input files
2) Small variation in c++ makefiles regarding the libraries
3) Cuba-3.0 updated to newest working version
12.05.2013
SVN rev 241
1) Reverted some Cuba updates to a stable version
2) Changed the integrator in paramloop.input to Divonne instead of Mathematica
3) Bug fixed in dotogetherC.pl when using "../launch" from a directory different from your standard working directory
29.04.2013
SVN rev 239
Warnings during subtraction and expansion in epsilon removed in form*.m which could occur in the case of tensor integrals and orders of epsilon in the numerator
17.04.2013
SVN rev 238
CUBA makefile altered to the version not containing the demos, new CUBA files added to subversion
17.04.2013
SVN rev 237
loop & general directory: Comment dimension statements like (*Dim=4-2*eps*) now also recognized as comments
16.04.2013
SVN rev 234
Update of the CUBA-library to the newest version
05.03.2013
SVN rev 233
1) Added demo files for the ggtt1 scalar diagram and the ggtt1 with contracted tensor numerator to demo directory.
2) Commented in some lines in the resultsloop.pl file which were commented out for testing.
28.02.2013
SVN rev 232
1) Small bugs in dotogetherC.pl removed concerning the usage of clusterflag=1, also the perlsrc/makesub.pm had to be altered for that.
25.02.2013
SVN rev 231
1) Application of symmetry relations to the numerator for externalegs=4 added to src/deco/calcFU.m.
2) Bug fixed concerning the renaming of epsilon to eps and the other way round in form*.m.
3) Implementation of togetherflag=1 for C++ version added. Perl script dotogetherC.pl steers the creation and compilation of integration files for togetherflag=1 using C++.
05.02.2013
SVN rev 230
1) Constraint s+t+u=Sum(p_i^2) removed from calcFU.m used to compute the functions F, U and the numerator
1) Comments in some files added, print statements removed
2) Small alteration of loop/demos/templateDBr1p3.m file for more user friendliness.
01.02.2013
SVN rev 229
1) Computation of numerator corrected in miscel.m
2) Computation of higher orders in epsilon occuring in numerator corrected and respective Perl scripts adjusted.
3) Factoring out of an overall prefactor occurring in the denominator of the numerator implemented.
23.01.2013
SVN rev 228
1) loop+general directory:
Resultfile output altered that an integration which had too few samples (CUBA Divonne output=-999) is translated into "probability that error is underestimated=1"
For that the makeint*.pm files needed to be changed.
2) Comments added to loop/src/deco/calcFU.m
22.01.2013
SVN rev 227
1) loop directory:
Option of having direct powers of epsilon in the numerator included. This occurs with integrals of rank >0 or may occur in user defined functions. For that, all form*.m, *.pl and (N)SDroutines.m files were changed accordingly. The maximal and minimal degree in epsilon in the numerator are written to the {graph}OUT.info file. From there, the minpole and required order in epsilon from which up to which the Laurent series is expanded, is adjusted.
2) Small bugs in formContourC.m removed where two semicolons were missing.
3) loop+general directory:
Resultsfile output slightly altered to improve user friendliness in case the integration files are incomplete or the integration did not have enough samples.
09.01.2013
SVN rev 226
1) Cuba library updated to Version Cuba-3.0 (date Jan 9 2013) where bugs in Divonne were fixed
07.12.2012
SVN rev 225
1) Minor changes to how the error probability is written to file. If fail is negative, a -999 is written to output file which translates as "integration failed, or not enough samples"
2) Names of *general.pl files and src/deco/templatetail_general.m were changed into *userdefined.pl and src/deco/templatetail_userdefined.m. Also the -n option is changed into the ./launch -u option to induce least confusion in the user. The -u in './launch -u' stands for _u_serdefined. The respective files were added to the SVN, the no longer needed files were removed from the SVN. The perlsrc/ files were changed accordingly. The readmes and the documentation are updated.
3) The parameters when using Divonne or Vegas were adjusted to more optimal ones in the paramloop.input and paramuserdefined.input file.
4) A bug occurring with the new Mathematica version 9 was fixed in the subandexpand*.m file allowing for the evaluation of integrands with arbitrary exponents.
5) Change of input primseclist={} to the more user friendly input name functionlist={} in the mytemplate.m file.
05.12.2012
SVN rev 224
1) loop directory: In form(P)*.m some n-1 were not replaced by feynpars which produced errors when n=feynpars.
2) loop&general directory: The error probability was sometimes not shown correctly in the output files when using other integrators than Divonne which is now fixed
2) general directory: correct choice of integrator implemented when integrating over less than 2 Feynman parameters
05.12.2012
SVN rev 223
1) Correction of makeint.pm file in general directory done and writing error probabilities to result outfiles also done now.
2) General directory:
Correction of some files which still had the CUBA-2.1 path in it. Alteration of results.pl and other files accordingly to write the error probability to the resultfile.
3) Minor changes to formatting of makeint.pm in loop directory and fixing of a small bug where a character was missing.
4) The outdated signcheck written to the graph.m file via the decomposegeneral.pl is removed to evade unnecessary error statements.
30.11.2012
SVN rev 222
1) Decreased size of sobol.cc and improved sobol.hh. Changed makemakeC.pm file such that sobol is not recompiled every time the user wants to compile a function. This decreases compilation time per function drastically.
2) install file altered. With the SecDec installation, sobol random sampler file is compiled as well. If compilation didn't work, an error message is echoed.
30.11.2012
SVN rev 221
1) Removed superfluous common block line in formfortran.m
2) Changed evaluation of "bigger" function to absolute values, so that with the two error probabilities of -999 and 0, -999 is chosen for the maxErrorProb value.
3) Commented out the eval statement in results*.pl files when collecting the contourdef=true results. Eval evaluates errors smaller 1.e-14 to zero. This is evaded if the eval statements are commented out.
4) Changed Cuba-3.0/src/divonne/Integrate.c file so that presampling of some points is not followed by a segmentation fault afterwards
22.11.2012
SVN rev 220
1) Changed computation of different possible lambdas for smalldefs=1 in makeintC.pm
2) Added hash "optlamevals=" to param*.input file to give the expert user the option to de-/or increase the number of evaluations done to find an optimal lambda when choosing "contourdef=true". This is good in cases of few functions and complicated diagrams.
21.11.2012
SVN rev 219
1) Small changes to makeint.pm to remove compiler complaints
2) Optimization -O to C++ compilation added in makemakeC.pm
3) Small changes to results*.pl file towards more user friendly {point}result{ord}.log files.
20.11.2012
1) Changed makeintC.pm and results*.pl such that in the C++ (w/o contour deformation) version the integration is done for a real part only and the results are collected accordingly.
2) results*.pl: When computing the prefactor with Mathematica, check if a Mathematica licence could be acquired and if not, there are 15 retries.
Integration resultfiles which contain no result for an imaginary part although they should (contourdef=true) are collected and redone, as are the integrations where the resultfiles are absent completely.
19.11.2012
1) Set precision in C++ result files to 10 digits. Now, the Fortran and C++ have the same precision.
2) Write error probabilities properly into all result outfiles and then the maximal error probability for each epsorder into the final resultfile.
3) Implemented the rescaling when using Fortran (before only for C++ available).
4) Changed C++ integration file: When computation of real part has finished, the result is already written to the resultfile *.out for each integration.
5) Added "cpp" (in addition to "Cpp") as allowed input for "language=" in myparamfile.input
15.11.2012
1) Shared definition feynpars in all form*P*.m files
14.11.2012
1) makeint.pm and getinfo.pm changed such that the accuracy achieved during integration is correctly translated into the result file
13.11.2012
1) Files justnumerics*.pl and finishnumerics*.pl updated for usage of togetherflag with Fortran and BASES
2) Integration of constants (independent of any Feynman parameter) possible now when using Fortran
3) Bug fixed related to collection of results when using launch and the -n option
4) Readme for noprimary option inserted
02.11.2012
1) Bug fixed for functions with more than 1 Feynpar but less than number of propagators-1 (Nn-1)
2) Print statement in subandexpand.m corrected
3) Print statements in makeint(C).pm changed so that resultfiles contain "errorprob=" instead of "chisq=" as chisq might be a misleading term
26.10.2012
1) NSDroutines.m and deconoprimary.m changed such that the numeration of the functions in the primseclist in the template file is just useful for the user and puts no constraints on the usage of the -n option. However, the first entry of every entry in the primseclist must remain a number, multiples of one number are allowed.
25.10.2012
1) Together-Flag fixed for Fortran BASES version
19.10.2012
1) Number of Feynman paramters the integrands depend upon is now written to infofile in every polestructure directory *l*h*/. Number is then read by polenumerics.pl and handed over to preparenumerics.pl to choose correct integrator (Divonne and Cuhre can't do 1d integrations) or an alternative integration setup (no Feynman parameters in function f*.cc) and correct number of Feynman parameters to be integrated in each epsorder.
2) Small changes to demo files paramloop.input and demos/paramA61.input
16.10.2012
1) Minor changes to demo files to make them work properly.
2) Changes to calcFU.m where ambiguous iterator name was used
3) Introduction of sdiQnoprim.m which does the same job as sdiQ.m, but for the general setup within the loop branch. The deconoprimary.m was altered accordingly.
15.10.2012
1) Changes to makeintC.pm for the case where no numerical integration is necessary (number of feynman parameters at integration stage is zero)
2) More comments added for clarity to templategeneral.m
3) Made variables Nloops, Dim and Nn known to prefactor.pl so that they can be used for computing the prefactor correctly
10.10.2012
1) makesum.pl for Fortran version changed to new naming conventions
2) Default values set for rescaleflag and cutconstruct
3) removed xl from decomposition step and changed into loops as global
variable (otherwise replacements as xl=loops=2 occur)
4) changed formindlist.m to previous version as previous one was
correct
04.10.2012
1) Fixed bug when user wants to compute loop integral with just 1 sector
2) Fixed bug when using Fortran or C++ in case of no contour deformation
3) Included sdiQnoprim.m to make predecomposition of squared Feynman parameters possible also for the "no primary sector decomposition" option
01.10.2012
1) Revised version of the README
2) Corrected the summation of results in the general case
3) Onshell conditions are now also imposed on the numerator, in the cutconstruct=0 case (here, numerator can be unequal 1)
4) In form*.m integrands such as 0^eps are recognized as zero integrals
5) The prefactor is altered such that all the rescaling is done on Mathematica Level until the writing of the *.f or *.cc files. After that, no rescaling issues need to be taken into consideration. The biggest invariant is additionally written to Monte Carlo integration files via Perl scripts.
27.09.2012
1) Fixed small bug in formContour(P)C.m for IBP=1, ";" was missing
2) Changed default input file for general files to paramgeneral.input and templategeneral.input
3) Wrote some comments to templategeneral file to get the syntax right
4) In results*.pl files rescaling is only done with the overall prefactor, not with the maximal invariant anymore. The rescaling with the invariant is done on the Mathematica level now by writing maxinv^(expoU+expoF) into the numerator
20.09.2012
1) Fixed small spelling mistake in in NSDroutines.m
2) Fixed rescaleflag. Rescaling is now done in Mathematica and no longer on Perl level. The decomposed U and F are divided by Symbol maxinv and then contour deformation and subtraction etc. is done. maxinv is written into Fortran/C++ files as "bi" (can also be changed back to maxinv).
28.6.2012:
gudrun
1) in header.pm: line 139: replaced $el/$r by eval($el)/$r to cope with case where $el=a/b
2) in makemakerun.pm: if(-e "/proc/cpuinfo"){..} enabled because this file doesn't exist on MAC
3) in makemakeC.pm: -lgsl -lgslcblas eliminated because it is not found on MAC
4) in install script: eliminated "installation successful" because it is printed even if Cuba compilation fails
5) eliminated compilation of default_tools in Cuba-3.0/makefile.in because it lead to compilation errors on some systems
15.03.2012:
1) Fixing bugs (endpointflag)
2) Optimizing memory use and time in Mathematica
3) Add more comments to integration outputfile (e.g. add total lambda used, including the lrs[][] value)
4) Change results.pl so that files are written to correct folder
09.03.2012:
1) When using IBP with CD files are written in between so that Mathematica won't run out of memory
2) Variable transformation implemented for integrands using CD which is done just before the MC integration
3) Makefiles compiling c++ functions improved
4) Writing of .cc functions instead of .c functions and changing compiler to gcc instead of g++
5) Fixing small bugs
14.02.2012:
1) results.pl got more lines of code in trade for a faster calculation of the prefactor.
In some cases Mathematica must not be used and in these cases Mathematica is not used anymore.
2) The numerical output file contains more comments now and the user can choose a smaller lambda,
even if the signs are different for the lambda and lambda^3 terms.
07.02.2012:
1) The file cleanup.pl now also cleans up the files created when using a cluster.
These start with "runtests-" and contain integration output, errors and cluster logs.
2) IBP is now possible with Contour Deformation, for that formContourC.m was changed.
3) Due to huge slowdowns the imaginary integrand now again starts with 1.0 instead of 0.0,
as there are more cases where the imaginary part is zero than cases where the imaginary part is 1.
11.01.2012:
modified justnumerics.pl and results.pl to enable multinumerics.pl to first make the compilation on the local
machine and then submit the jobs
modified the param.input file so that it now contains the new options
09.12.2011:
put stable deformation into loop folder
10.11.2011:
modified: src/subexp/ExpOpt.m
src/subexp/formContourC.m
perlsrc/prefactor.pl
perlsrc/header.pm
perlsrc/makeintC.pm
perlsrc/preparenumerics.pl
results.pl
prefactor is calculated for rescaled invariants, functions to find the biggest invariant and rescale all invariants is written into the header; these functions are called in makeintC.pm to write the rescaled invariants into the intfile.c files and they are called in results.pl in order to get the correct prefactor;
09.11.2011:
modified:
perlsrc/makeintC.pm
results.pl
src/subexp/ExpOpt.m
src/subexp/formContourC.m
makeintC.pm: small changes in how the intfile.c + intfile.h file is written
ExpOpt.m: small changes in how the fi.c files are written
formContourC.m: predefinition of lambda in mathematica file commented out
26.10.2011:
modified:
loopdevelop/perlsrc/makeintC.pm
loopdevelop/perlsrc/makemakeC.pm
loopdevelop/perlsrc/polenumerics.pl
loopdevelop/perlsrc/preparenumerics.pl
loopdevelop/src/subexp/ExpOpt.m
loopdevelop/src/subexp/formContourC.m
makeintC.pm writes a header file for the .c-files, a new function is added to the (point)intfile(kk).c determining the biggest invariant and dividing every invariant by the biggest one in case it is >1.
makemakeC.pm is changed accordingly to allow a smooth compilation of all .c files including the header in the f(i).c functions before writing the makefile.
polenumerics.pl and preparenumerics.pl got small comments
ExpOpt.m got a new module which writes a small infofile with the exponent of F in it.
formContourC.m got changes related to the writing of the f(i).c files. Additionally to the writing of the infofile, a new function was introduced which has not yet any meaning and the header of the f(i).c functions was changed.
14.10.2011:
Starting values for lambda, alpha, beta can be written into the parameter file. These or standard starting values are written to subandexpand*l*h*.m
11.10.2011:
added new folder loopdevelop in which the contourdeformation will be developed
18.8.2011:
in Mathematica 8, MatchQ[x_,0] does not match if x=0. (real instead of integer),
therefore MatchQ had to be redefined (decomposition.m changed) and == has been replaced by
MatchQ in primarySD.m lines 76/77.
11.08.2011:
- start implementing contour deformation
needs to be added: (not done yet)
loop/src/subexp/formContour.m (to replace formfortran.m in contour def case)
modified:
loop/src/subexp/subandexpand.m
loop/subexp.pl
loop/perlsrc/header.pm
loop/param.input
25.07.2011
- further changes for C-version.
added:
loop/perlsrc/makeintC.pm
loop/perlsrc/makemakeC.pm
modified:
loop/src/subexp/formC.m
subandexpand.m
ExpOpt.m
/util/Format.m
loop/perlsrc/makemakerun.pm
makesub.pm
mathlaunch.pm
getinfo.pm
polenumerics.pl
preparenumerics.pl
writefiles.pm
batch
loop/results.pl
subexp.pl
justnumerics.pl
loop C version functional for Euclidean kinematics, and in physical region for
1-scale problems.
11.07.2011
- to start implementing a C-version in addition to the Fortran version,
added:loop/src/subexp/formC.m,
modified:
param.input,
header.pm
subexp.pl
subandexpand.m
ExpOpt.m
makeint.pm
preparenumerics.pl (for name of intfile)
16.06.2011
-functions can be left implicit for the algebraic part of the calculation for
the general program.
-more control over the random number generator seed is now possible.
-documentation brought up to date for new functionality.
02.06.2011
-added files permitting to loop over different values for numerical constants
in "general" subdirectory
01.06.2011
-allowed length of output filenames increased in general program
26.05.2011
-unregulated singularities detected for general program
14.04.2011
-problem with collection of lowest pole in loop program fixed
11.04.2011
-version number changed to 1.1.0 to indicate inclusion of Cuba library as significant upgrade
-detection of singularities within the integration region refined
04.04.2011
-previous problems with restarting the program after manual interrupt fixed.
24.03.2011
-development branch added
22.03.2011
-in the general program constant prefactors not dependent upon epsilon were not
treated correctly after the update of 03.03.2011. This is fixed
03.03.2011
-exponents of the form (a_i+b_i eps) are now accepted with b_i implicit for the
general program (a_i must of course be explicitly given as it dictates what
level of subtraction is needed). The Hyp4F3 example is changed to give a
demonstration of this
01.03.2011
-routine findsymmetries in loop/src/subexp/formindlist.m improved
-justnumerics.pl updated to remain compatible with previous change in batch
files
23.02.2011
-SDroutines.m. more thoroughly documented
-Other robodoc headers brought up to date
-'grouping' option added to param.input files
-submission of Mathematica jobs via batch*l*h* retrys when no
licences are immediately available
-informative print statements adjusted for the loop program when
sectors are decomposed seperately
21.02.2011
-preferred set selection during the decomposition was incorrectly
implemented in the loop case, this is fixed.
-Portable Batch System output/error files renamed;
they now appear in the directory the job submission file is written in.
16.02.2011
-common block 'ndim' in fortran files caused a naming conflict. This is now
called 'ndimen'
08.02.2011: release of version 1.0.1
Changes with respect to version 1.0.0:
08.02.2011
- Bug fixed relating to prefactors and collection of the results in the loop directory.
Version 1.0.0 gave an incorrect result when "togetherflag=1" and
"primarysectors" and "multiplicities" were non-empty.
Specifically if k sectors are in the list "primarysectors", the
stated result was k times too large.
- general/cleanall.sh and loop/cleanall.sh added to run any launchclean scripts
remaining, and remove all intermediate files from the main directory (e.g. job
output/error files from the batch system)
04.02.2011
- New sector decomposition routines added to allow for parallelisation of the
Mathematica part
- Cuba library 2.1 integrated into the package and into the user interface
6.12.2010
- issue with "togetherflag=1" on 64bit machines fixed
- "graphname" starting with a number is now also possible
- info about timings and result files added to the "demos" directories
1.12.2010:
- general/finishnumerics.pl: fixed option to call the script from any directory
- general/demos/param5F4.input: changed parameter settings to have a