ChangeLog.md

GM2Calc-2.2.0 [July, 31 2023]

New

• The Mathematica interface can now be disabled at the CMake level with the CMake option ENABLE_MATHEMATICA. The Mathematica interface is enabled by default.

  Example:

  cmake -DENABLE_MATHEMATICA=OFF ..
  

• The Python interface can now be disabled at the CMake level with the CMake option ENABLE_PYTHON. The Python interface is enabled by default.

  Example:

  cmake -DENABLE_PYTHON=OFF ..
  

• The tests can now be disabled at the CMake level with the CMake option ENABLE_TESTS. The tests are enabled by default.

  Example:

  cmake -DENABLE_TESTS=OFF ..
  

Changes

• Change (commits 2e0a7b0d, 80bb314e): Performance improvement of DR-bar to on-shell conversion for parameter points where the conversion does not converge.

• Change (commit 1ca5d76b): Change detection of Eigen and Boost headers in CMakeLists.txt to be compatible with Conan 2. See README.md for installtion instructions.

Fixed bugs

• Bugfix (commit ebce8648): Fix compilation error when boost and Eigen headers are located in different parent directories. Thanks to Sho Iwamoto.

• Bugfix (commit 577b33c4): Fix numeric divergences in fermionic 2-loop contributions in the 2HDM.

GM2Calc-2.1.0 [March, 18 2022]

Changes

• Change (commit fd9ec530): Performance improvement of the 2-loop electroweak contributions in the THDM by a factor 4.

• Change (commit 00728f28): Performance improvement of the 2-loop contributions in the MSSM.

Fixed bugs

• Bugfix (commit d2d12dc2): Correcting the calculation of alpha_h in the 2HDM when the Higgs boson mass eigenstates h and H change their positions in the multiplet due to mixing effects. This bug has been seen in a scenario with tan(beta) < 1 and sin(beta - alpha_h) ~ 1.

GM2Calc-2.0.0 [October, 25 2021]

New features

• Calculation of the anomalous magnetic moment of the muon in the Two-Higgs-Doublet Model (2HDM) at the 1- and leading 2-loop level. The implemented expressions are taken from [JHEP 01 (2017) 007]. See the README file and examples/example-thdm.{c,cpp,m} for examples.

Fixed bugs

• Bugfix (commit 205f5cf): Ensure that MSSM Goldstone bosons are always at the first position in the Higgs multiplets. This change has no effect on the value of a_mu in the MSSM.

• Bugfix (commit b72cc64): Avoid a NaN in the 2-loop sfermionic MSSM contributions when the f_sferm function is called with z = 0 as argument. This case appears when a stop, sbottom or stau is massless.

• Bugfix (commit 906bbca): Avoid a NaN in the 2-loop sfermionic MSSM contributions when the f_S function is called with z = 0 as argument. This case appears when a chargino is massless.

• Bugfix (commit 5bee954): Avoid a NaN in the 2-loop sfermionic MSSM contributions when the f_PS function is called with z = 0 as argument. This case appears when a chargino is massless.

GM2Calc-1.7.5 [February, 11 2021]

Fixed bugs

• Bugfix (commits 6f83231, 3fcc914): Fix compilation on Visual Studio 15.

GM2Calc-1.7.4 [February, 07 2021]

Fixed bugs

• Bugfix: Fix linking for debug build on Linux with Clang.

GM2Calc-1.7.3 [February, 07 2021]

Fixed bugs

• Bugfix: Fix compilation on Windows with MSVC 16 (2019).

GM2Calc-1.7.2 [January, 10 2021]

Fixed bugs

• Bugfix (commit 65b196b): Fix compilation on 32-bit ARM, i.e. Raspberry Pi.

GM2Calc-1.7.1 [October, 02 2020]

Changes

• Change (commit 04cfaeb): Update dilogarithm to cleaner and more performant version from the polylogarithm package version 5.0.0.

• Change (commit 4e65a64): Update FindMathematica to version 3.3.0

GM2Calc-1.7.0 [April, 26 2020]

Changes

• Change [commit 62bf8d6]: Update dilogarithm to more performant version from the polylogarithm package version 3.4.0. Results in a performance improvement of GM2Calc of up to 15%.

GM2Calc-1.6.0 [March, 23 2020]

New features

• Feature: New cmake/FindGM2Calc.cmake file.

Fixed bugs

• Bugfix (commit c06f713): The function find_right_like_smuon() has been corrected, which finds the most right-like smuon. This bugfix affects only the SLHA input (the GM2Calc-specific input is not affected). On parameter points that are affected, a change of the value of amu by up to 90% has been found.

  We kindly thank Ipsita Saha and Sven Heinemeyer for a detailed analysis of this bugfix.

• Bugfix (commit 5e020ee): In case of SLHA input, the neutralino mixing matrix is now read from the NMIX block and used to determine the bino-like neutralino. This change fixes convergence problems when converting the DR-bar input parameters to the on-shell scheme.

Changes

• Change: Separate public from private headers. The public headers are located in include/gm2calc/, while the private headers are located nearby the respective source files. The public headers must be included as

  #include "gm2calc/gm2_version.h"
  

• Change: Private header and source files renamed:

  • Model-specific source files are named as MSSMNoFV_onshell*.{h,hpp,cpp}.

  • Files specific to the calculation of (g-2) are named as gm2_*.{h,hpp,cpp}.

• Change (commit f276703): In case the SLHA input contains multiple HMIX blocks, use the last one to define the renormalization scale. Input parameters are then searched for in blocks with that renormalization scale.

• Change (commit 9907e86): The Mathematica interface now returns Indeterminate for amu in case there is a problem with the calculation.

• Change (commit aad392c): Update FindMathematica to version 3.2.7

• Change (commit 059744f): Performance improvement by avoiding frequent re-calculation of mb(MZ,DR-bar).

• Change (commit b38eab7): Update of default fine-structure constant from PDG (2019) and [arXiv:1802.02995].

• Change (commit 67b39a5): New error message when tan(beta) is too large, resulting in floating point overflows.

• Change (commit 9fdc594, e73c63a): Performance improvement of reading SLHA input.

• Change (commits 4ad9d95, 445ea1c, c7a5e9d): More stingent test of GM2Calc configuration options. Bail out if invalid options are given.

• Change (commits ea03df9, 32df289, 6325c07, 438b421): Performance improvement of conversion of SLHA input to GM2Calc-specific on-shell scheme.

• Change (commits d54714f, cf8e702, ec60044, da8d948): Make GM2Calc compatible to be used as CMake sub-project.

• Change (commit 77615d7): Verbose, error and warning messages are now written to cerr. This allows to separate the physics output from the informational messages.

• Change (commits ee1eab3, 0635231, 576afcb): Avoid redundant error messages to stdout/stderr. This significantly improves the performance of parameter scans using the Mathematica interface, where frequent flushing of the stdout/stderr buffers is expensive.

• Change: Many stylistic internal improvements; fixes of clang-tidy warnings.

GM2Calc-1.5.2 [Feature, 08 2020]

• Change: Updated FindMathematica to version 3.2.6, which can detect the Wolfram Engine.

GM2Calc-1.5.1 [January, 21 2020]

• Feature: New Mathematica file math/ffunctions.m with analytic expressions for the loop functions.

• Feature: New Mathematica file math/amu2Lapprox.m with analytic expressions for the 1- and 2-loop fit formulas.

• Change: Improved performance of 1-loop functions.

GM2Calc-1.5.0 [May, 11 2019]

• Feature: New make install target, so GM2Calc can be installed. The installation includes the executable gm2calc.x, the GM2Calc library, the public headers and a corresponding gm2calc.pc file for pkg-config.

  Example:

  cmake -DCMAKE_INSTALL_PREFIX=${HOME}/.local/ ..
  make
  make install
  

• Feature: The required Boost and Eigen libraries can now be installed with Conan:

  mkdir -p build
  cd build
  
  # install dependencies
  conan install ..
  
  # invoke cmake as usual
  cmake ..
  

• Change (commit 577815e): gm2calc::Error now inherits from std::runtime_error.

• Change (commit 674cc13): If there is more than one entry with the same key in an SLHA block, use the last one.

• Bugfix (commit 2ea83e8): Catch unphysical input parameter where MW >= MZ.

GM2Calc-1.4.3 [October, 07 2018]

• Bugfix (commit cd218cd): Properly initialize trilinear A parameters in copy constructor of MSSMNoFV_onshell.
  Note: This copy constructor is never used in GM2Calc.

GM2Calc-1.4.2 [July, 14 2018]

• Bugfix (commit e784837c): Fix cmake error from FindDoxygen.cmake when building with cmake < 3.3, by enabling policy CMP0057. Thanks to Sho Iwamoto.

GM2Calc-1.4.1 [June, 13 2018]

• Change (commit 2114aef): The default output format of the command line program gm2calc.x has been changed: If no GM2CalcConfig[0] entry is provided, then the output is

  • written to the GM2CalcOutput block for SLHA input
  • written in detailed form to stdout for GM2Calc input

• Bugfix (commit 304d771): Catch non-numeric SLHA input. Thanks to Peter Athron and the GAMBIT collaboration.

GM2Calc-1.4.0 [March, 27 2018]

• Feature: Replace GNU make build system by cmake to improve platform independence. See the README.md file for build instructions.

• Bugfix (commit 8d0bac6): Reame quad() function to fix a compilation error on Windows/Cygwin.

GM2Calc-1.3.3 [July, 19 2017]

• Optimization (commit aa7afc0): Avoid redundant calls to complicated lambda_mu_cha() function.

• Bugfix (commit cbb4df2): Fix compilation error on Cygwin where M_PI might not be defined in <math.h>.

• Bugfix (commits 9d4f768, d820a53, b9a6320, 54970a6): Catch floating point overflow/underflow from diagonalization of mass matrices during DR-bar to on-shell conversion of right-handed smuon mass parameter.

• Bugfix (commit 54b86e8): A small coefficient in the complex dilog has been corrected. However, the complex dilog is not used so far.

GM2Calc-1.3.2 [February, 22 2017]

• Bugfix (commit e8af0c9): Catch potential exceptions from C interface functions which calculate amu w/o tan(beta) resummation.

• Bugfix (commit 654276e): Allow user to pass numeric values to GM2CalcAmuSLHAScheme[], GM2CalcAmuGM2CalcScheme[] and GM2CalcSetSMParameters[] which are not numbers, but which would evaluate to numbers.

GM2Calc-1.3.1 [January, 31 2017]

• Bugfix (commit ab73c49): Workaround bug in mcc 11.0.0.

GM2Calc-1.3.0 [July, 21 2016]

• Feature: A Mathematica interface for GM2Calc has been added. To build it, run make mathlink. The compiled MathLink executable can then be found in bin/gm2calc.mx. To use it, the MathLink executable must be installed in Mathematica via

  Install["bin/gm2calc.mx"];
  

  Two examples using the Mathematica interface of GM2Calc can be found in

  examples/example-gm2calc.m
  examples/example-slha.m
  

  These two examples behave exactly like their C/C++ counterparts.

• Change (commit b3c7357): Abort calculation if tan(beta) is undefined or zero.

• Change (commit 910b1bd): Abort calculation if the mu parameter is zero. For mu = 0 the approximate fermion/sfermion 2-loop corrections are ill-defined.

• Change (commit 16b181b): Abort calculation if the lightest chargino mass is zero. In that case the photonic 2-loop contribution originating from loop function F3C is ill-defined.

• Change (commits 6d2a8b3, 2a8b2ed): Create shared library src/libgm2calc.so (in addition to the static library) for convenience.

• Change (commit c0b5923): Compile all .cpp files in src/ and put all generated object files into the libraries, except for src/gm2calc.o. In this way, there is no need for GAMBIT to modify src/module.mk to add further .cpp files for compilation.

• Bugfix (commit b3d3c0f): Implementation of the x = 0 limits of the F1N[x], F2N[x], F3N[x], F4N[x] and F1C[x] functions.

• Bugfix (commit 3ec5df4): Catch potential exception during iteration to determine the QCD scale, Lambda_QCD. This exception is thrown, for example, if the Z pole mass is set to be larger than 5 TeV.

GM2Calc-1.2.0 [June, 21 2016]

• Feature (commit b65d75d): Adding C interface functions to retrieving warnings and problems in form of C strings.

• Feature (commit ffe9b50): Perform the calculation of the uncertainty of a_mu(0-loop) and a_mu(1-loop). The uncertainty of a_mu(0-loop) is the magnitude of a_mu(1-loop). The uncertainty of a_mu(1-loop) is the sum of the magnitudes of a_mu(2-loop,best) and the uncertainty of a_mu(2-loop,best).

• Change (commit fbee2e3): Adding calculation of amu uncertainty to the C/C++ example programs.

• Bugfix (commit 0665a09): Correcting the used sbottom masses in Delta_b corrections.

• Bugfix (commit 71b0cf3): Do not allow the calculation of amu for negative soft-breaking squared sfermion mass parameters.

• Bugfix (commit 7d4d4e7): Use better initial guess for the root-finding algorithm, which determines the soft-breaking squared mass parameter of the right-handed smuon from the mostly right-handed smuon pole mass.

• Bugfix (commit 9eecf61): Catching potential exception during the DR-bar to on-shell conversion of the soft-breaking squared mass parameter of the right-handed smuon.

GM2Calc-1.1.2 [April, 08 2016]

• Bugfix (commit a374bf7): Reformulate Delta_b corrections to avoid numerical problems when Mu, M3 or M1 are zero.

• Bugfix (commit 4177460): Implement limits of Fa(), Fb() and amuBmuLmuR() functions when one or more masses go to zero.

GM2Calc-1.1.1 [March, 29 2016]

• Bugfix (commit 9ca7890, 33e68a4, c047d46): Implement limit of Iabc() function when one or more masses go to zero.

GM2Calc-1.1.0 [December, 14 2015]

• Change (commit 5cbbc50): The example programs have been moved to a separate examples/ directory.

• Change (commit 873133f): Adding a C interface. The new headers

  src/gm2_1loop.h
  src/gm2_2loop.h
  src/gm2_uncertainty.h
  src/MSSMNoFV_onshell.h
  

  declare C interface functions for GM2Calc routines. In addition, two example C programs

  examples/example-gm2calc_c.c
  examples/example-slha_c.c
  

  have been added to illustrate the C interface. These two C example programs behave exactly like their C++ counterparts. They can be compiled by running make examples.

GM2Calc-1.0.0 [October, 29 2015]

• Official release 1.0.0.

GM2Calc-0.2.17 [October, 26 2015]

• Feature (commit 132851): The uncertainty of amu(2L,TB resummed) can be calculated by setting GM2CalcConfig[5] to 1. Depending on the chosen output format (GM2CalcConfig[0]) the uncertainty is written

  • as a single number to stdout in case of minimal output,
  • to the first line in case of detailed output,
  • to GM2CalcOutput[1] in case of NMSSMTools output,
  • to GM2CalcOutput[1] in case of SPheno output,
  • to GM2CalcOutput[1] in case of GM2Calc output.

• Feature (commit 02841d): A new SLHA-compliant output format GM2Calc has been added to avoid interference with SPheno or NMSSMTools: If GM2CalcConfig[0] is set to 4, the value of amu is written to GM2CalcOutput[0]. If uncertainty estimation has been enabled in addition, the uncertainty is written to GM2CalcOutput[1].

• Feature (commits 25eee8d, 148adaf): Adding two C++ interface examples for the SLHA-compliant and the GM2Calc-specific input format. The examples can be found in the files src/example-slha.cpp and src/example-gm2calc.cpp.

• Change (commit 3bf585a): Memorize all tachyonic particles (not only one).

• Change (commit af21077): Print warning if tachyons exist in detailed output without tan(beta) resummation.

• Change (commits 1426497, f558f22): Reduce number of digits after the decimal point to 8 digits in the printed result for amu and its uncertainty.

• Bugfix (commit 4b3f634): Check for tachyons if tan(beta) resummation is disabled.

GM2Calc-0.2.16 [October, 05 2015]

• Change (commit 7a4f62a): The example input files have been renamed such that their file name extension reflects their input format. input/slha.in -> input/example.slha input/gm2calc.in -> input/example.gm2

• Bugfix (commits e784353, c194f9c): Fix compilation problem with clang++ on Mac due to different STL implementation of std::conj(). Thanks to Björn Sarrazin.

GM2Calc-0.2.15 [October, 01 2015]

• Bugfix (commit 585614f): Be less strict in checking if scales of two SLHA blocks are the same, as SPheno might write the value of the scale with different numerical precision to the block header. Thanks to Björn Sarrazin.

GM2Calc-0.2.14 [September, 30 2015]

• Bugfix (commit 53187a9): Fix compilation error with g++ 4.6.3.

GM2Calc-0.2.13 [September, 30 2015]

• Change (commit 9230209): Print neutralino and chargino mixing matrices in verbose output.

• Change (commit 810c56b): Use mb(MZ) in the DR-bar scheme instead of mb(mb) in the MS-bar scheme.

GM2Calc-0.2.12 [September, 27 2015]

• Change (commit 6205c0b): The strong gauge coupling g3 is now set to the non-zero PDG default value of 0.1184.

• Change (commit 45b914c): The C++ user interface for changing the values of vu and vd has been simplified: Now the user only needs to set TB. The VEV v = sqrt(vu^2 + vd^2) is calculated internally using the W and Z pole masses.

• Change (commit 0659005): Regenerate default SLHA input file (CMSSM 10.1.1, [arxiv:1109.3859]) with FlexibleSUSY 1.2.2. Note: slightly updated SM input parameters are used compared to SoftSUSY's CMSSM 10.1.1 version inOutFiles/lesHouchesInput .

• Change (commit 5e993dd): 2-loop fermion/sfermion contributions from 1st and 2nd generation sleptons have been added to Delta_g1 and Delta_g2 (Eqs. (6.6a)-(6.6b) [arxiv:1311.1775]). Patch: Dominik Stöckinger.

GM2Calc-0.2.11 [September, 24 2015]

• Change (commit 16a8118): Don't throw an exception if gluino mass, M3, is zero, as it contributes at the 2-loop level.

• Change (commit 53cbee3): Always print model parameters in verbose mode. In addition, in verbose mode the DR-bar to on-shell conversion iteration steps are printed.

• Change (commit 5854532): Replace Fortran implementation of Li2(z) by C++ implementation. The C++ implementation is faster and eliminates the need of a Fortran compiler.

• Change (commit d31daa0): Make use of LAPACK library optional. By default the Eigen library is used for diagonalization of mass matrices.

• Change (commit f105385): If SLHA input format has been chosen and no GM2CalcConfig input block is provided, the default output format will be SLHA and the value of amu will be written to the SPheno bock SPhenoLowEnergy[21].

• Change (commit 3f6d5db): If the conversion of the right-handed soft-breaking smuon mass parameter from the DR-bar to OS scheme fails using a FPI, a root finding algorithm is tried.

• Bugfix (commit 2f481f5): The sorting of the smuon pole masses got lost during the FPI, which lead to non-convergence of the right-handed soft-breaking smuon mass parameter with FPI.

• Bugfix (commits 47c8237, bd15f57, 64cc024): Read DR-bar parameters from SLHA blocks with the same scale. As renormalization scale the renormalization scale of the HMIX block is chosen.

GM2Calc-0.2.10 [September, 02 2015]

• Bugfix (commit d4baa95): fix compilation with g++ 4.7.4

• Feature (commit 8993e26): New flag GM2CalcConfig[4] to enable/disable verbose output.

• Change: Nicer formatting of detailed output.

GM2Calc-0.2.9 [September, 01 2015]

• Bugfix (commit 4a5725a): Use default muon mass if no muon mass is given in the SMINPUTS block.

• Bugfix (commit c12dfe3): Refine and implement limits for the functions F*C, F*N, Fa, Fb and Iabc. Patch: Markus Bach.

• Bugfix (commit d3cc2de): Determine soft-breaking left-handed smuon mass parameter from the muon sneutrino pole mass.

• Change (commit 6fc7ed4): add 2-loop 2L(a) corrections to best approximation for amu. Patch: Markus Bach.

GM2Calc-0.2.8 [August, 06 2015]

• Change (commit 2260ac4): If a physical problem has occured, the problem description is added to SPINFO[4] in case of SLHA-compliant output.

• Feature (commit bfc22cc): Allow user to force output even if physical problem has been spotted.

GM2Calc-0.2.7 [August, 06 2015]

• Bugfix (commit c514eca): Implementation of the limit x=1 for the functions F3C, F4C, F3N, F4N.

GM2Calc-0.2.6 [August, 06 2015]

• Bugfix (commit e16898f): Implementation of the limit x=1 for the functions F1C, F2C, F1N, F2N. Patch: Markus Bach.

GM2Calc-0.2.5 [August, 04 2015]

• Bugfix (commit 8acb0b7): Make loop order selectable in case of minimal output.

• Change: Do not print any output if the spectrum contains a tachyon.

GM2Calc-0.2.4 [August, 04 2015]

• Bugfix: Implementation of functions Fa(x,y), Fb(x,y) and H2(x,y) in the limit x=1 and y=1.

• Bugfix: Implementation of I(a,b,c) in the limit of equal arguments.