GoSam is hosted by Hepforge, IPPP Durham

General questions

How can I install GoSam?

What is GoSam-contrib?

GoSam uses various reduction and integral libraries (Ninja, Golem95, Samurai, OneLoop etc.). They are bundled in a separate package called gosam-contrib.

FORM and QGraf are not contained in the gosam-contrib, as they are often pre-installed.

If not present, all the required components are installed automatically by the GoSam install script.

I have found a bug. How can I report it?

Please use the ticket option on the Wiki (login with "guest","gosam") or report to gosam@projects.hepforge.org

Process generation

I tried to rerun GoSam on a process. Why do some files still have the old content?

Files like matrix/test.f90, Makefile.conf, common/config.f90, common/model.f90 are often modified by the user. This is the reason why they are not overwritten. Add the flag -z to GoSam to recreate all files.

I get the FORM error message No memory while allocating WorkSpace?

You need to reduce the size of the heap that FORM allocates. Use form.workspace=100 in your input card (or another smaller value than the default 1000 which is for systems with >= 2GB RAM).

You can also change this afterwards: Call find . -name "form.set" | xargs sed -i "s/WorkSpace.*/WorkSpace 100M/" in the process directory or edit the form.set files by hand.

I get the FORM error message Program terminating in thread 0 at RemoveNCContainer Line 9?

The FORM 4.1 binaries on the FORM homepage do not match the original 4.1 release. They include recent changes, which introduced bugs and incompatibilities with GoSam.

Please use the original 4.1 FORM version from https://github.com/vermaseren/form/releases/tag/v4.1-20131025

You can see the latest changes in FORM and reported issues at https://github.com/vermaseren/form

I get compiler error messages similar to contract2_4 is not defined?

GoSam supports fully currently only one additional tensor rank in the numerator,
i.e. rank-3 bubbles, but not rank-4 bubbles.
For up to 4-point functions, Golem95 could be extended easily to (nearly) arbitrary ranks.
Only files of its tensorial reconstruction interface need to be re-generated.
Please contact the authors.


I get the error message `Error: There is no specific subroutine for the generic 'cond_t'`?

This error messages appears if you try to compile code generated by GoSam with a different floating point precision in common/config.f90 than the libraries in gosam-contrib were built with. For more details see the related question

Can I use GoSam with quadruple precision?

Note GoSam has advanced methods to check the result for a phase space point and recompute it with a different method if necessary. If an unstable point is detected, GoSam uses its rescue system (normally Golem95C). Hence, there should be no need to switch to quadruple floating point precision in general. The settings for the rescue can be controlled by the parameters PSP_check, PSP_rescue, PSP_chk_th1, PSP_chk_th2, PSP_chk_th3, PSP_chk_kfactor in the input card.

As in some cases it could still be interesting to calculate results for specific phase space points in higher precision, you can re-compile the whole code to enable quadruple precision. For this, you need to comment/uncomment the corresponding lines in common/config.f90. Then you need to call make clean and make compile to rebuild the whole code.

Please note that before, you also have to re-compile all libraries which GoSam uses (bundled in gosam-contrib) in quadruple precision. See the related question.

You may need adapt the compiler flags in Makefile.conf such that the gosam-contrib with quadruple-precision is used. Secondly, you need to source the correct gosam_setup_env.sh into your shell or set LD_LIBRARY_PATH accordingly.

What is BLHA or BLHA2?

The Binoth-Les Houches-Accord Interface is an interface between one-loop-providers like GoSam and Monte Carlo event generators.

For the most recent BLHA2 standard see http://arxiv.org/abs/1308.3462 . The original BLHA standard is described in http://arxiv.org/abs/1001.1307 .

How can I update GoSam or its components?

Run the install script again without any options in the folder where your install-log.ini file is located.

How can I uninstall GoSam?

Run the install script in the folder where your install-log.ini file is located with the -u flag.

During the installation I have selected the option to use a local version of FORM or qgraf. How can I install their version coming with GoSam?

Run the install script in the folder where the install-log.ini file is located and append form and qgraf to the command line.

Is there an offline version of the installer?

No, not yet. Feel free to send a patch to us which modifies the installer so that it can be used offline.

How can I install gosam-contrib in quadruple precision?

The most easy option is to install GoSam a second time and appending --quadruple to the installer command line.

If you are using a manual installation, you need to append --with-precision=quadruple to the ./configure script of gosam-contrib.