FeynHiggs: History of the Code

Last update: Oct 06, 2022
What happened (FeynHiggs2.19.0):

• Complete reimplementation of fixed-order two-loop corrections (based on arXiv:1401.8275, arXiv:1409.1687, arXiv:1705.07909, arXiv:1802.09886) improving the numerical stability, using results from arXiv:1910.02094.
• Support for DRbar renormalization of sbottom sector in fixed-order calculation.
• The bottom/sbottom corrections now contain the full phase dependence without any approximation/interpolation.
• Support for MDRbar renormalization of stop/sbottom sector drastically improving precision for m_gluino >> m_stop (based on arXiv:1912.10002).
• SLHA interface extended to provide BLOCKs necessary for the usage of HiggsBounds/HiggsSignals.
• Fixed bug in the calculation of the A/h₃ boson width if using the ifort compiler (thanks go to Keith Olive).

• Fixed bug in M_W calculation (relevant for mass-degenerate sleptons). This bugfix is preliminary, as it may lose a few MeV in accuracy. Thanks go to E. Bagnaschi.
• Fixed bug in EWPO calculation in the case that SLHA was used as input. Thanks go to E. Bagnaschi for testing.
• Updated default values for AlfaHad and invAlfa0 to 2020 PDG numbers.
• Fixed bug in the selection of complex THDM EFT calculation. This bug led to problems for loglevel = 4 (shifts of O(200 MeV) in M_h, completely erronerous values for the other neutral Higgs masses).
• Corrected signs of imaginary Yukawa RGEs for complex THDM EFT calculation. The numerical effect on M_h is negligible.
• Fixed SM input parameters for low tanβ LHCHWG benchmark scenarios (in the example folder).
• Fixed compiler issue with Ubuntu 20.

• Implemented complex THDM as EFT (see arXiv:2010.04711 ). This EFT calculation is automatically activated for loglevel = 4 and tlCplxApprox > 0. It is also used if the EWino mass scale is larger than MSusy. In all other cases, the real THDM-EFT calculation from arXiv:1805.00867 is employed.
• Implemented the O(α_t²) threshold corrections to the Higgs self-couplings for the (real and complex) THDM as EFT (see arXiv:2010.01989 ).

• The top-Yukawa coupling is now by default extracted from the OS top-quark mass at the 3L level. This leads to downwards shifts in M_h of ~ 0.5 GeV, see arXiv:2009.07572 for more details.
• Implementation of a link to the code Himalaya for the calculation of O(α_t α_s²) threshold correction, see arXiv:1807.03509 and arXiv:2009.07572 for more details. See the README for instructions on how to activate Himalaya (by default Himalaya is not run).
• Implementation of the full phase dependence for the SM-EFT calculation (loglevel = 1,2,3). See arXiv:2009.07572 for more details.
• Bug fixed in the function FHGetSelfgl (it gave back the full self-energies before).
• Bug fixed in the O(α_t²) threshold corrections, numerical impact in M_h of O(10 MeV).
• Bug fixed in the renormalization of the one-loop self-energies (negligible numerical impact).
• Bug fixed in uncertainty estimate of EFT calculation (negligible numerical impact).

• Consistently added stop-sector CKM rotation (even in the case of all NMFV-δ = 0) to all NMFV squarks. This improves (via the M_W² counterterm) the decoupling behavior of Γ(h → ff) for large squark masses and has a (mostly small) numerical impact on the EWPO calculation.
• Bug fixed in non-logarithmic subtraction terms relevant if the flags looplevel != 0 and loglevel != 0. The numerical effect on M_h is negligible. Thanks go to Ivan Sobolev for noticing the bug.

• Implemented 2HDM as EFT (activated by setting loglevel = 4). See arXiv:1805.00867 for more details. Recommended if M_A << M_SUSY.
• tanβ = TB is now by default defined as tanbMSSM(M_SUSY) (was tanbMSSM(scalefactor*MT) before).
  tanbTHDM can be used by setting environment variable FHTBDEF = 1. Non-default renormalization scale can be set by using environment variable FHTBSCALE.
• μ = MUE is now by default defined as MUE(M_SUSY) (was MUE(scalefactor*MT) before).
  Non-default renormalization scale can be set by using environment variable FHMUESCALE.
• Added example scripts to run LHCHXSWG MhEFT125 and MhEFT125chi scenarios, see arXiv:1901.05933 for details.
• Bug fixed for DRbar stop input parameters (leads to shifts of O(0.1 GeV) in M_h).

• Improved uncertainty estimate in the Higgs-boson mass calculation (see arXiv:1912.04199 for more details).
  The uncertainty estimate works only fully correctly for the best available calculations! (This implies in particular loglevel = 3, looplevel = 2.)
• Changed parametrization of non-logarithmic subtraction terms. This leads to changes in M_h of O(0.1 GeV) (see arXiv:1912.04199 for more details).
• Bug fix in finite field renormalization in case of complex input parameters; this can lead to changes in M_h1 O(1 GeV) for large phases and multi TeV scalar quark masses.
• The flag botResum = 2 is the new default.

• Implemented two-loop Δ_b corrections from arXiv:1001.1935 (thanks to Michael Spira for providing the full expressions). This correction is activated by setting the flag botresum = 2. While the two-loop corrections are always included in the calculation of the Higgs decay widths, they enter the calculation of the Higgs masses only for looplevel = 1. This restriction to looplevel = 1 is a temporary feature that will be lifted once the implementation of the corrections from arXiv:1705.07909 is completed. Thanks go to Ivan Sobolev.
• Fixed bug in the calculation of MSbar vev leading to changes in the SM-like Higgs mass of O(0.1) GeV.
• Fixed bug for the flag runningMT = 1 leading to changes in the SM-like Higgs mass of O(0.1) GeV.
• Fixed bug in case of DRbar sfermion input parameters leading to changes in the SM-like Higgs mass of O(0.5) GeV for large A_b. Thanks go to Ivan Sobolev.

• Higgs propagator poles are now (again) determined by using a numerical pole finding procedure (as in FH2.13.0 and before), however, with a change in the field renormalization, see the next item. This change improves the behavior in the vicinity of level crossing. (Thanks go to Pietro Slavich, Stefan Liebler and Tim Stefaniak for relentless testing.). Numerically, this leads to shifts of O(100 MeV) for most parameter choices (for low tanβ and/or M_A larger shifts are possible).
• To avoid the problems concerning the inclusion of uncontrolled higher-order logarithms (via the iterative pole finding procedure, see above) found in arXiv:1706.00346 a finite field renormalization has been introduced (in contrast to a pure DRbar renormalization used so far). It was ensured that the definition of ZHiggs and tanβ in our output (which in principle could be affected by the choice of the Higgs field renormalization) has not changed.
• Calculation of h_i → γγ has been augmented: SM limit reach in the case of strongly mixed neutral Higgs bosons is improved.
• A bug in the evaluation of m_b including resummed higher-order corrections has been removed; the numerical effect is (very) small. Thanks go to Ivan Sobolev.
• Numerical stability in the non-degenerate (α_t²) threshold corrections in the EFT calculation has been improved.

• Eigenvalue ordering during determination of pole masses changed from "ascending" to "aligned with tree-level eigenstates (h,H,A)". Should improve behavior in the vicinity of level crossing. (Thanks go to Pietro Slavich, Stefan Liebler and Tim Stefaniak for relentless testing.)
• In the EFT calculation for complex parameters the O(α_tα_s) threshold corrections are now interpolated correctly for complex M₃.
• For DRbar stop/sbottom input: In the h_i → h_j h_k calculation the stop vertex loops are now calculated with DRbar stop parameters.
• For botResum = 0 the bottom Yukawa coupling is now evaluated via m_b(m_t) · (1 – Δ_b) (was m_b(m_b) before).
  
• Numerical stability of O(α_tα_s) and O(α_t²) corrections to the Higgs-boson self-energies in the cMSSM improved.
• The erroneous combination of MHp as input and the flag tlCplxApprox = 0 is now explicitely forbidden.
• Bug in the one-loop H^+- field renormalization counterterm in the gaugeless limit has been fixed. This can lead to very small changes in the Higgs-boson masses and ZHiggs.
• Message reporting fixed for Windows version of MFeynHiggs (was broken since gfortran4.x).
• New configure option "--real10" for x86 extended precision (between double and quad, check ./configure --help).
• configure changed: Cygwin build improved.

• In the EFT calculation the O(α_t²) threshold corrections for the non-degenerate mass case have been implemented (taken from arXiv:1703.08166 [hep-ph]).
• In the EFT calculation a bug in 1L electroweakino threshold correction has been fixed; the numerical effect on M_h is smaller than O(100 MeV).
• The fit function for σ(gg → t H^+- b) has been improved (instabilities for tanβ < 6 have been removed).
• Bug in the DRbar-OS conversion for sbottom masses used for EWPO fixed.
• In the case of DRbar input for stops and sbottoms:
  The stop masses used for all calculations but the Higgs boson masses are now again OS masses (were DRbar masses in FH2.14.0).
  The output for Stop and Sbottom masses are again always OS masses.

• Two flags have been removed: fieldren and tanbren are now fixed at their default values (0 and 0), see FHSetFlags.
  Consequently, FeynHiggs is now called with 8 flags (instead of 10 before).
• (Optional) DRbar renormalization of the scalar top sector (i.e. of m_stop1,2 and X_t) has been included. The renormalization scale is choosen to be Q_t (if loglevel != 0, Q_t must be set equal to the stop mass scale).
  Note: in this case the output of stop (sbottom) parameters is DRbar (OS).
• A new option for the determination of pole masses (and ZHiggs) has been included, avoiding the problems found in arXiv:1706.00346 [hep-ph].
  The old version can be activated by setting the environment flag FHFOPOLEEQ to 0.
  The effect can be a downward shift of about 300 MeV in M_h, and of about 200 MeV in δM_h.
• The EFT result for loglevel > 0 is now obtained using the MSbar vev and not the OS vev as before
  (v_OS² * λ(m_t) → v_MS² * λ(m_t)).
• A modified two-loop LL subtraction term for loglevel > 0 has been introduced to better capture the logarithmic behaviour in case of runningMT = 1,2. This can lead to an upward shift in M_h of about 1 GeV.
• The gluino threshold now works also together with loglevel = 3.
• The mass counterterm for the A-boson changed at the two-loop level: it now contains additionally the term
  Im[Σ_AA^(1)'(m_A²)] Im[Σ_AA⁽¹⁾(m_A²)].
  This ensures a constant A-boson mass when renormalized OS (and used as input parameter).
  For complex parameters and M_H^+- as input, the two-loop mass counterterm of the charged Higgs is changed accordingly.

• looplevel = 0, 1 can now consistently be called with loglevel = 0, 2, 3 (we are working on 1). For high SUSY mass scales, where numerical instabilities can occur in the Feynman-diagrammatic calculation, those corrections can be switched off at the users's will (looplevel = 0 corresponds to a pure EFT calculation). Still experimental!
• For loglevel > 1 the one-loop and O(α_tα_s) threshold corrections are now more complete (extension to the case of non-degenerate soft SUSY-breaking parameters, inclusion of missing term in the matching conditions of the effective Higgs-Higgsino-EWino couplings). The numerical effect on M_h is mostly small, O(100 MeV).
  Thanks to Pietro Slavich for providing the general expression for the O(α_tα_s) threshold correction.
• Interpolation of the EFT calculation in the case of complex parameters included.
• The flag tlCplxApprox now has only the (previously) allowed values 0, 1, 2, 3. The choice of the parameters for the complex interpolation, previously done by setting tlCplxApprox = 3, 4, 5, 6 is now done by calling FHSelectIpol or by setting the environment variable FHSELECTIPOL.
• Higgs production cross sections for WBF, ZH, WH, ttH at 13 TeV included and for 14 TeV updated. SM cross sections are taken from the LHCHXSWG. (ttH@14 TeV still missing.)
• The electroweak precision observables have been improved and reorganized. Thanks go to Lisa Zeune!
  • The EWPO are now evaluated calling the new routine FHEWPO.
  • Δr is now evaluated in the SM and MSSM. In the SM it is given via a Fit formula from hep-ph/0608099. In the MSSM it is based on arXiv:1311.1663 [hep-ph] (by L. Zeune et al.) It contains a full one-loop calculation in the cMSSM and two-loop corrections to Δρ interpolated for the complex parameters.
  • The Δρ output now contains the one loop contribution from the third generation of (s)quarks, the gluon and gluino two-loop corrections (and the one from the sbottom mass shift).
• Bug in the call of the flavor observables removed.

Several bugs for runningMT = 3 removed. They also affected in general the scalefactor variation.
Thanks go to Pietro Slavich for relentless testing!

• Log resummation now works also for m_t^OS.
  Therefore, the estimate of the uncertainty from higher order corrections in the top/stop sector obtained by switching between m_t^OS and m_t^MSbar(NNLO) also works consistently in the case of log resummation.
  The uncertainty evaluation only works reliable for the flag choice resulting in the highest precision! In particular, this means mssmpart = 4, looplevel = 2, loglevel = 3, runningMT = 1 .
• Small bug-fix in the logarithmic subtraction terms. They are now consistently evaluated for y_t^MSbar = m_t^MSbar/v^MSbar. For loglevel > 0 this yields a small shift in M_h.
• DRbar-OS transition (for stop/sbottom sector input) improved:
  The reconstruction of the DRbar mass matrix now employs m_t^DRbar(Q). The DRbar-OS shifts now use α_s(Q).
• Bug in the scale variation Q → Q/2, 2 Q in the uncertainty calculation removed. The scale variation is not also applied in the two-loop rMSSM corrections.
• Bug in the resummation routines for loglevel=3 fixed. It was numerically relevant only for |X_t^OS/M_SUSY| > 2.5. Thanks go to Peter Athron for reporting the bug.

• Log resummation via RGE running improved, now controlled with the new flag loglevel.
  • loglevel = 0: no log resummation
  • loglevel = 1: one high SUSY mass scale (corresponds to looplevel = 3 in previous versions)
  • loglevel = 2: three high SUSY mass scales are taken into account correctly: M_stop, M_gluino, M_cha/neu. Electroweak contributions are taken into account in the RGE's and in the m_t^MSbar evaluation, leading to a downward shift of about 1 GeV.
  • loglevel = 3: the RGE resummation is performed at the 3-loop (NNLL) level; correspondingly the threshold effects are evaluated at the 2-loop level.
  The running top Yukawa coupling is now evaluated from the running vev.
• The RGE running now starts at m_t^pole, instead from m_t^MSbar as before.
• Small improvements in the NNLO evalution of m_t^MSbar, effect of about -60 MeV (changing from iterated to non-iterated evaluation).
  The evaluation of the Higgs mass uncertainty has changed accordingly.
• α_t(m_t) now evaluated from the (new) NNLO value of m_t^MSbar. This leads to a downward shift in M_h of a few hundred MeV.
• Bug in the charged Higgs cross section evaluation ("alt") fixed.
  For Sqrt[s] != 8 TeV the tanβ enhancement factor was missing.
• New in the default output: X_t, X_b.

• Implementation of the SM NNLO MSbar top-quark mass, to be used instead of the NLO value, controlled by the flag runningMT. The new options for this flag are
  
    runningMT = 0  :  top pole mass
    runningMT = 1  :  top SM MSbar mass at NNLO (new)
    runningMT = 2  :  top SM MSbar mass at NLO (previously 1)
    runningMT = 3  :  top MSSM DRbar mass at NLO (previously 2)
  
  Using the NNLO top-quark mass value effectively corresponds to a downward shift by 1.8 GeV. This NNLO value can be used together with the log-resummation in the top/stop sector, where it constitutes the matching value at the top-quark mass scale.
  
  
• Bug in the Xt OS-MSbar conversion in the log resummation fixed.
  Thanks go to H. Bahl.
• Production cross sections for charged Higgs bosons at 13 TeV included, for details see arXiv:1409.5615.
• Numerical instabilities in the O(α_t²) corrections to the Higgs boson self-energies removed.
• Updated version of the light stop benchmark scenario is provided.

• Bug in the two-loop corrections to MHp removed.
• FHGetSelf now delivers also unrenormalized self-energies and Tadpoles.
  Note: the syntax of the call has changed accordingly.
• runningMT = 2 can now be chosen for all values of tlCplxApprox.

• Implementation of a calculation of the Higgs-boson self-energies using a DRbar top-quark mass (consistently at the one- and two-loop level; not yet combinable with the log-resummation). This calculation is invoked by setting the flag
    runningMT = 2 ,
  but so far is only implemented for
    tlCplxApprox = 1 .
  Details can be found in arXiv:1505.03133.
  (Thanks go to Sophia Borowka for detailed testing.)
• Bug in O(α_tα_s) and O(α_t²) in the cMSSM removed: M3SD = M3SU had (incorrectly) been forced.

• Corrections of O(α_t²) to the Higgs-boson self-energies in the complex MSSM included, for details see arXiv:1409.1687 [hep-ph]. They are included automatically when tlCplxApprox != 0.
  These corrections are also included into the 2-loop corrections to the charged Higgs boson mass in the real MSSM, for detail see arXiv:1502.02394 [hep-ph].
  For M_A or M_H⁺ the O(α_tα_s) and O(α_t²) corrections are now included also with tlCplxApprox = 0.
  (Thanks to Sebastian Paßehr for relentless testing!)
• Bug in FHAddSelf removed. (Thanks go to David Kunz.)

• Change in the calculation of Γ(H → hh): contributions beyond one-loop from the top/stop sector have been included in the vertex corrections (to match the corrections included in the full calculation of M_h).

• In the loop-induced decays Φ → γγ, γZ, gg as well as in σ(gg → Φ) the bottom Yukawa couplings, containing the resummed Δ_b corrections, have been improved to show full decoupling to the SM for large SUSY mass scales. (Thanks go to Motoi Endo.)
• Higgs decays that are not evaluated at the full one-loop level are now calculated with the UHiggs matrix instead of the ZHiggs matrix as before. This ensures the decoupling for the heavy neutral Higgs bosons. More details can be found in [FHSelectUZ].
• The maximum length for parameter tables has been increased, in particular to allow to read in the low-tb-high scenario.

• The full one-loop correction from Lepton Flavor Violation (LFV) have been included for the Higgs-boson self-energies and the electroweak precision observables.
  They are activated by setting (at least) one of the LFV δ's to a non-zero value, see [FHSetLFV] for details.
  More information about the implementation into FeynHiggs (as well as into FeynArts/FormCalc) can be found in arXiv:1408.0663 [hep-ph].
• Bug removed in the theory uncertainty evaluation for the third Higgs-boson mass in the presence of complex parameters (thanks go to Sebastian Passehr).
• Bug in FHAddSelf when used within Mathematica removed (was there since FH2.10.1).
• In the DRbar-OS conversion of the squark input parameter now the SU(2) is enforced in the OS parameters (by enforcing a universal M3SQ as obtained from the stop sector conversion).
• Fixed compilation issues for Mathematica 10.

• The leading momentum dependent two-loop correction at O(α_tα_s) have been included, see arXiv:1404.7074 [hep-ph].
  These new corrections make use of the program SecDec.
  For usage of this new feature take a look at the README.

• The evaluation of Higgs boson masses (in particular M_h), mixing angles and field renormalization constants has been improved for large scalar top masses (in the multi-TeV range). A resummation to all orders of leading and subleading logarithms of the type log(m_stop/m_t) has been included, see arXiv:1312.4937 [hep-ph] for details. The new calculation is activated by setting the flag looplevel = 3.
  Important: for very large values of m_stop above 10 TeV and negligible mixing in the stop sector quadruple precision might be needed to get a reliable result.
• LHC Cross section for pp → H^- t^* + X at 8 TeV included. So far it is only available (roughly) for tanβ > 30. (Thanks go to Martin Flechl.)
  To get the cross section for H^- and H⁺ production, a factor of 2 has to be applied.
  For this cross section, please cite:
  • T. Plehn et al. hep-ph/0206121; see also here
  • S. Dittmaier, M. Kramer, M. Spira, M. Walser, arXiv:0906.2648
  • M. Flechl, M Kramer, R. Klees and M. Ubiali in arXiv:1307.1347.
• Bug for complex parameters in interpolated(!) two-loop corrections removed. Was in there since FH286.
• Decays of H/A to scalar neutrinos included.
• The output for the sbottom-like squark masses now contains the Δ_b corrections in the off-diagonal entry. Internally nothing has changed in the calculations.
• MathLink compilation fixed for MacOS10.9.

• The decay mode Higgs → Z γ is finally included (SM and MSSM).
• Conversion of A_b from DRbar to OS has been corrected to match exactly the implemented two-loop corrections. This is implemented so far at O(α_s), where the remaining terms should be marginal. (Thanks go to P. Slavich!)
• Removed overcounting of running m_t in the Δ_b entering the two-loop Higgs self-energy calculation. (Thanks go to P. Slavich.)
• In the presence of NMFV effects, a separate set of MFV sfermion masses and mixings (with off-diagonal elements neglected) is now used in the MFV-only parts of FeynHiggs, e.g. all two-loop corrections, rather than the corresponding entries of the full NMFV matrices. This guarantees unitarity and is furthermore unaffected by level crossings, which could distort MFV-only results considerably.
  Note that the syntax of the FHGetPara API call has changed since it now returns both sets of sfermion masses and mixings.
• The gluonic two-loop contribution to Δρ is now evaluated over all sfermion generations.
• Bug fixed in the 2-loop (g-2)_μ calculation.
• The new benchmark scenarios as given in arXiv:1302.7033 [hep-ph] are now included as default scenarios. In example/LHBMS/ the scenarios can be found in complete agreement with the publication can be found. In example/LHCHXSWG/ the scenarios as used by the LHC Higgs Cross Section Working Group can be found (including a script to create the corresponding SLHA files).

• Bug removed that caused faulty calculations for runningMT = 1 and looplevel = 1 (thanks go to Oscar Stal and Leo Galeta).
  Bug was present in FH2.9.[0-3].
• Calculation of M_W improved: now the Higgs mass with the largest coupling to WW is used for the SM evaluation.

• Bug in the MathLink fixed (was present only in FH292).
• Redefinition: Key → FHKey in CFeynHiggs.h and MFeynHiggs to be compliant with ROOT.
• Better SLHA1 compliance in SLHA output (relevant when FH is the first step in an SLHA chain).

• Bug in the 7 TeV cross section removed. In intermediate versions the result was given in 1/pb, not in 1/fb.
• Man pages updated.
• Minor fixes for the C/C++ interface.

• Bug removed that resulted in a small shift in the charged Higgs mass when it was chosen as input value and p2approx = 4 was used.
• Quad-precision now also works for MFeynHiggs.
• Enlarged SLHA(1) output in case FeynHiggs is used as the first step in an SLHA chain; results in improved readability of subsequent codes.
• Significantly restructured C/C++ interface (CFeynHiggs.h). Works with C99 and interfaces to the quadruple-precision version using the long double type in C (for details see 1107.4379).

• Improved h_i → γ γ decay rates which show the correct decoupling behavior. (Largest effect came from the different treatment of the charged Goldstone boson loop in the SM and MSSM.) Thanks go to Oscar Stal for relentless testing.
• 7, 8 TeV cross sections added for VBF, WH, ZH, ttH.
  14 TeV cross sections for those channels update from here.
  SM cross sections are taken from the LHC Higgs cross section working group.
  MSSM correction factors are applies as for 14 TeV.
• Potential problem in the DRbar-OS transition fixed.
  The problem occurred for large (μ tanβ). In this case the correction to A_b could become very large, unreliable and jeopardizing the whole evaluation of Higgs masses etc. Now in case the correction turns out to be larger than 30% the DRbar-OS transition for the scalar bottom sector is switched off. Consequently, also the involved two-loop corrections to the Higgs boson self-energies, O(α_sα_b), O(α_tα_b), O(α_b²) are switched off.
• Bug in the evaluation of m_t(Q) removed. If runningMT = 0 was chosen, the running started (wrongly) at m_t^pole, and not at m_t(m_t). For runningMT = 1 no error was made.
• The choice of the flag runningMT now only affects the top mass used in the Higgs boson self-energy correction, but not enter anymore in, for instance, the Higgs decay evaluations.
  Similarly, the flag botResum only affects the bottom/sbottom contribution in the Higgs boson self-energies (however, in other calculations where appropriate Δ_b is used as before, just the flag has no impact anymore).
  More specifically: The flags affect mainly the Higgs boson mass evaluation, but not the subsequent evaluation of decay widths, cross sections etc. This can lead to smaller changes in some observables for some specific choice of flags.
  These changes apply similarly to all three generations of squarks.
  The sbottom mass output (also via FHGetPara) now corresponds to m_b(m_b) (without any Δ_b correction.
  A summary of the various calculations of Yukawa couplings, squark masses etc. can be found here.
• New output call for sbottom sector parameters that are used in O(α_sα_b), O(α_tα_b), O(α_b²) Higgs boson self-energy corrections: [FHGetTLPara] (see the man page).
• Now all sfermion masses are part of the screen output.
• Bug in the Δρ (and thus in the M_W and sin²θ_eff) removed (was in since FH28?). This bug was effective only for very large sfermion masses. Thanks go to Pietro Slavich.
• Charged Higgs decays are now evaluated for V_CKM =! 1.
  The Δ_b-type corrections to the bottom-type Yukawa couplings are now also included in generation mixing decays, e.g. in Γ(H^+- → bc).
  The HpSfSf macros have an additional generation index now.
• In the case of MFV the LL-entries in the stop-type mass matrices agree now exactly with the sbottom-type matrices, i.e. the rotation with V_CKM that is required in the case of NMFV is not executed in the MFV case.
• Bug fixed in the NMFV calculation of Higgs boson self-energies. Thanks go to Arnd Behring and Stefan Schacht.
• Name conventions for the NMFV parameters changed slightly.
  Also LFV parameters processed (but not yet actively used in any FeynHiggs computation).
• Default value for p2approx was changed from 0 to 4, i.e. now by default UHiggs is evaluated at p² = 0, see [FHSetFlags].
• The configure option --quad now also works with gfortran (possible since gfortran 4.6).
• Minor fixes in the SLHA reader (possible sign reading error in NMIX, ...)
• Still some problems with quad precision in MFeynHiggs persist; will be solved for FH291.

• Reparametrization of running top mass in the two-loop terms improved. The difference between the two options runningMT = 0, 1 is of pure three-loop order. Small effects on masses and uncertainties for runningMT = 1.
• Decay H -> VV* refitted for M_H < 161 GeV, minor changes.
• Problem in the SLHA reader fixed (was present only in FH2.8.5) that led to problems for complex parameters.

• Fixed problems in configure with SLC.
• Fixed inconsistent scale of alpha_s in MSbar → DRbar conversion (Thanks to Pietro Slavich).
• Fixed problem in SVD for all-zero diagonal.
• Fixed naming of H⁺ → f f' decays in the textual output.

• Γ(h_i → h_j h_k) now included with the full one-loop corrections, for details see arXiv:1103.1335 [hep-ph].
• Γ(h_i → ff) improved to match exactly arXiv:1103.1335 [hep-ph].
• Charged Higgs decays are now evaluated for CKM <> 1. New decays are H^- → cb etc.
  Also Δ_s,d are now included here for the decays within one generation (so far only Δ_b was used).
• The decays h_i → WW, ZZ are now evaluated with the SM results taken from Prophecy4f (at the NLO level; thanks go to Ansgar Denner, Ivica Puljak and Daniela Rebuzzi). The MSSM results are obtained, as before, by applying the effective coupling approximation.
• Bug in the O(m_t⁴) correction to M_H⁺ removed.
• Bug in the DRbar-OS conversion fixed: the stop/sbottom conversion routine was before accidentally also excecuted for staus.

• Bug in charged Higgs cross section removed. So far the cross section was evaluated for M_A, not for M_H⁺.
• Several Higgs cross sections got re-fitted, leading to differences below 1%.
• Problem with Higgs productions at very high Higgs boson masses removed (which could lead to a floating exception).
• All hard links were replaced by soft links to improve AFS compatibility.

• Bug in Γ(H⁺ → tb) removed that was active for M_H⁺ < m_t (an error in the interpolation; thanks go to Phillip Gutierrez).
• ``make all'' is no longer necessary. configure determines whether MathLink executables can be created. This is then done automatically with ``make''.
• Improved compilation with ifort.

• Changes in Γ(h_i → ff):
  
  • After the introduction of the full one-loop corrections to all Γ(h_i → ff) for extreme parameters it was possible that large negative one-loop corrections could drive the respective Γ < 0. This is now avoided by adding always the full loop-squared term.
    Of course, this problem occured only in extreme scenarios with a very small tree-level contribution.
  • In Γ(h_i → ff) now also self-energy type diagrams on the externa Higgs boson leg with a Higgs mixing with a Z boson or the Goldstone boson G are included.
  • In case of Γ(h_i → bb) the interference term at the one-loop level now also includes the Δ_b corrections.
  • The implementation of Γ(h_i → ff) now corresponds to arXiv:1103.1335 [hep-ph].
• Bug in the φ-b-b effective coupling has been removed.
• The Higgs boson mass limits in FHHiggsProd for which cross sections are evaluated are now inclusive for the limit values, not exclusive as before. Obviously this only has an effect if exactly the limit value was reached with one Higgs boson mass.
• A new function was implemented: FHGetNMFV: it gives back all derived NMFV parameters, i.e. ...
• The internal structure of the NMFV parameters has changed. This should have no visible effect for the user.
• A bug in the table function removed.
• Removed division-by-zero condition when setting CKM parameters to zero.
• Mathematica output now also contains the console output, e.g. references to other codes used within FeynHiggs.

• Improvements for the σ(gg → h_i) evaluation:
  • Tevatron: finer grid for c-factors.
  • Tevatron and LHC: Second evaluation for σ(gg → h_i) included: k-factors for top-contribution, bottom-contribution and top-bottom-interference contribution are used.
  • Tevatron and LHC: Updated SM NLO normalization cross section, based on MSTW2008, including finer grid around 2 m_t.
  (thanks go to A. Vicini for the data and to Stefan Liebler for testing).
• Improved Tevatron cross sections for σ(bb → h_i). SM-evaluation is now based on hep-ph/0304035 (by Harlander and Kilgore), using MSTW2008. The MSSM effective bbh_i coupling is evaluated as before. (Thanks go to Markus Warsinsky.)
• user-defined value of α_s is now used everywhere.
• Bug in evaluation of the complex Higgs-boson self-energies removed that was present in FH2.7.4, but not in earlier versions.
• Bug (typo) in the coupling of Higgses to scalar fermions appearing in the Higgs decays removed. This affects (nearly) all decay modes at the few per-cent level, depending on the parameters. This bug was presents since FH2.7.0.
• Bug in FHAddSelf removed.
• O(α_bα_s) corrections (thanks go to Pietro Slavich):
  Stop and sbottom masses in a sub-sub-leading two-loop contribution to the Higgs-boson self-energies are now evaluated correctly omitting the D-terms.
  Hard-wired value of m_b replaced by user-defined value m_b^MSbar,SM(Q=m_b).
  Hard-wired value of v is replaced by the one evaluated from G_F, also possibly user-defined.
• In the Higgs decays now m_t(M_hi) is used (instead of m_t(m_t) before).
• Arguments in the calls to several existing functions have changed (see the API guide for details):
  • FHSetSMPara, FHRetrieveSMPara: the new call for FHSetSMPara reads

            subroutine FHSetSMPara(error,
         &    invAlfaMZ_, AlfasMZ_, GF_,
         &    ME_, MU_, MD_, MM_, MC_, MS_, ML_, MB_,
         &    MW_, MZ_,
         &    CKMlambda_, CKMA_, CKMrhobar_, CKMetabar_)
    

    It should be noticed in particular that the last two CKM parameters are the bar-parameters.
    The same parameter structure holds now for FHRetrieveSMPara.
  • FHReadRecord: the new call reads

            subroutine FHReadRecord(error, record, slhadata, inputfile)
    

    FHReadRecord now reads native files as well as SLHA files.
  • FHSetSLHA.
  • FHSetRecord.
  • FHFlavour: Now included are BR(b → s γ), Δ M_Bs in NMFV (full one-loop calculation, no `mass insertion' method). The new call reads

           subroutine FHFlavour(error,
             bsgammaMSSM, bsgammaSM,
             deltaMsMSSM, deltaMsSM,
             bsmumuMSSM, bsmumuSM)
    

    The last two variables concerning BR(B_s → μ μ) are so far dummies and will be part of future extensions of FeynHiggs.
• Several new functions, mainly regarding input and output, have been included (see the API guide for details):
  • FHOutput: writes FeynHiggs inputs and outputs to a file.
  • FHOutputSLHA: writes FeynHiggs inputs and outputs to a file in SLHA format.
  • FHGetSMPara: gives back all derived SM parameters, i.e. the CKM matrix.
  • FHSLHARecord: it transforms SLHA input into the Record format.
  • FHRetrieveRecord: fills a FeynHiggs Record with the FeynHiggs parameters currently set.
  • FHRetrieveOSPara: it gives out the on-shell parameters, and can be used if the user input was given in the DRbar scheme.
    On the other hand, FHRetrievePara gives back exactly the user input (possibly DRbar, possibly on-shell).
  • FHSetFlagsString, FHRetrieveFlagsString: instead of calling FHSetFlags with all the flags as argument, one can now alternatively call FHSetFlagsString. Example:

            call FHSetFlagsString(error, "400202113")
    

    The retrieve function, FHRetrieveFlagsString, is also included.
• The output format can be defined also in the command line mode.
  The native format normally creates a screen output, but now can also create an SLHA file, example:

   FeynHiggs nativeformat.in#SLHA (flags)
  

  The SLHA format normally creates a new SLHA file, but now can also create screen output, example:

   FeynHiggs slha.in#FH (flags)
  

• The scale factor in the command line mode is no longer an additional parameter in the input line, but has to be included in the input file itself. The corresponding line should read (example):

   scalefactor         1
  

• Bugfix in the effective bb-Higgs coupling (only in the effective coupling, not in any production or decay mode).
• Finer grid for σ(gg → h_i) calculation (thanks go to A. Vicini).

• Bug fixed in the Γ(t → H⁺b) calculation.
  Before the Δ_b contributions were included twice.
  Bug was present only in FH2.7.[0,1,2].
• SU(2) compliance of NMFV improved.

• Bug fixed in the complex two-loop Higgs boson self-energy corrections.
  Changes occur only for the Higgs mass uncertainty calculation.
• Bug in the SLHA reader fixed. A-terms of first and second generation were read in wrongly.

• Two bug-fixes in FHAddSelf and DRbartoOS.

• New "amplitude corrections factors" for gg →φ (φ = h,H,A) to match better the SM calculation. Thanks go to A. Vicini
• New subroutine FHAddSelf.
  FHAddSelf allows the user to include/add (momentum independent!) shifts in the Higgs-boson self-energies, to be used consistently in the computation of the Higgs masses and mixings.
  It has to be called after all SM and MSSM parameters are set.
  A flag controls whether the contributions to the CP-even self-energies are given before (φ₁-φ₂ basis) or after (h-H basis) the rotation with α.
  Take a look at the corresponding man pages.
• Bug fixed in Γ(t → H⁺ b). This bug was present only in FH270.
• Bug fixed in Γ(H⁺ → t b). This bug was present only in FH270.

• The LHC Higgs production cross section σ(gg → h) has been improved.
  • In the bottom Yukawa coupling we use m_b(m_b). This can lead to a strong increase in the cross section for large tanβ.
  • The renormalization cross section for the SM is now calculated only up to NLO. This ensures that the NNLO corrections are only applied to the top loops.
  • The SM prediction is now taken from Grazzini et al., see that web page for an extensive list of references.
• Bug in the calculation of ttH/A has been fixed. Changes in tth should be minor.
• Bug in the calculation of M_H+ in for the choice mssmpart = 1 has been fixed (thanks go to Tomas Gonzalo).

• The full one-loop contributions to Γ(Φ → f₁f₂) have been included. This includes also flavor-violating decays (f₁ ≠ f₂).
  Thanks go to Karina Williams for checking and providing results. Thanks go to Dominik Scherer for providing data on Δ_b.
  Call to FHCouplings slightly modified. The "excl" option has been superseded by HiggsBounds. The new last argument "fast" specifies whether the off-diagonal Higgs to fermion decays are evaluated.

    fast = 0 : all Higgs to fermion decays are evaluated
    fast = 1 : only the diagonal Higgs to fermion decays are evaluated.
  

• The LHC Higgs production cross sections (FHHiggsProcs) is now envoked with the collider energy as an argument. Some production cross sections are now available for various center of mass energies.
  The SM bb-Higgs production is taken from hep-ph/0304035 by R. Harlander and B. Killgore using the respective code bbh@nnlo.
  More details can be found in the man page of [FHHiggsProd].
• Phase interpolation in the two-loop contributions improved.
  Now the flag tlCplxApprox can take the values from 0 to 6:
  

    #9 = tlCplxApprox, complex parameters in the 2-loop corrections
      0: all corrections evaluated in the rMSSM
      1: a_s a_t in cMSSM only
      2: a_s a_t in cMSSM, rest in rMSSM
      3: a_s a_t in cMSSM, rest interpolated in At,Ab,M_3,MUE (default)
      4: a_s a_t in cMSSM, rest interpolated in At,Xb,M_3,MUE
      5: a_s a_t in cMSSM, rest interpolated in Xt,Ab,M_3,MUE
      6: a_s a_t in cMSSM, rest interpolated in Xt,Xb,M_3,MUE
  

  If phase interpolation is wanted, the user has to specify what his input variables are. The interpolation is then performed in exactly these variables.
• the flag runningMT now only has two options:

    #7 = runningMT: determines which top mass shall be used in the corrections:
      0: MTpole directly and in the stop masses
      1: MTrun directly and in the stop masses
  

  More information can be found in the change-log from Dec. 20. 2008.
• Calculation of uncertainties related to Δ_b improved. Now the uncertainties are evaluated as the difference in the Higgs masses for Δ_b in the denominator (resummed) and in the numerator (not resummed).
• External Z-factor for charged Higgs included to ensure the on-shell properties for an external charged Higgs.
• Call to set the NMFV parameters has changed:

          subroutine FHSetNMFV(error,
       &    deltaLL12_, deltaLL23_, deltaLL13_,
       &    deltaLRuc_, deltaLRct_, deltaLRut_,
       &    deltaRLuc_, deltaRLct_, deltaRLut_,
       &    deltaRRuc_, deltaRRct_, deltaRRut_,
       &    deltaLRds_, deltaLRsb_, deltaLRdb_,
       &    deltaRLds_, deltaRLsb_, deltaRLdb_,
       &    deltaRRds_, deltaRRsb_, deltaRRdb_)
  

  It is now taken into account automatically that the LL entries have to be equal for top- and bottom-type squarks (modulo CKM rotation that is also taken into account).
• Bug in the calculation of M_H+ in the NMFV case fixed.
• Bug in the settings of Sbottom masses in the complex 2L corrections of O(α_tα_s) fixed.
  This bug could lead to a large error for M_gluino >> M_SUSY.
• Calculation of ΔM_s inclucded (also for NMFV, added in collaboration with J. Illana).
• Γ(H⁺ → tb) has been fixed for M_H⁺ < M_W + 2 m_b. Now it is (correctly) set to zero. Thanks go to Pietro Slavich.
• Error in the NMFV input parameters corrected. Before no NMFV parameters were recognized. This problem appeared only in FH265.
• FeynHiggs can read input from stdin, use "-" as file name.

• Use of running top mass changed, flag tl_runningMT changed to runningMT. Now the same top mass is used in the one- and two-loop corrections (and in the evaluation of the scalar top masses).
  • runningMT = 0: the pole mass is used everywhere.
  • runningMT = 1: the running mass is used everywhere, except in the stop mass matrix for the evaluation of the stop masses and mixing angle. The same stop masses and mixing angle are used for one- and two-loop corrections.
  • runningMT = 2: the running mass is used only for the evaluation of the stop masses and mixing angle.
  • runningMT = 3: the running mass is used everywhere (default).
  
  The running mass is defined as the SM MSbar mass at the scale of the top mass. The corrections included into the SM MSbar top mass are consistent with the choice of the two-loop corrections. If only O(α_tα_s) corrections are included at the two-loop level, only the O(α_s) term is included in the SM MSbar top mass. If in addidtion also the O(α_t²) corrections are included at the two-loop level, also the O(α_t) term is included in the SM MSbar top mass.
  This means especially that using only the one-loop corrections (flag looplevel = 1) the flag runningMT has no effect.
  The theory uncertainties derived from the change in the top mass are evaluated as the difference of runningMT = 0 and 1 (2 and 3) if runningMT = 0 or 1 (2 or 3) has been chosen by the user.
• Scale of α_s entering the Δ_b corrections changed from Q_old = m_t to Q_new = Sqrt[m_sbot1 m_sbot2].
  However, the induced change should be considered as an additional theoretical uncertainty.
• The Δ_b corrections in the one-loop result for the Higgs boson self-energies are now only included if the respective extra two-loop self-energies O(α_bα_s) are included. Otherwise the running bottom mass, m_b(m_b) is used for the one-loop self-energies. Thanks go to Karina Williams.
• DRbar-OS transition for m_b corrected. In FH264/263 the implementation lead to uncertainties in the interpolation routines in the sub-leading two-loop corrections in the presence of complex phases.
• Bug in the DRbar-OS transition for scalar tops and bottoms removed. The bug was effective when the complex O(α_t α_s) routines were used.
• new k-factors for gg → Φ implemented. Thanks go to Alessandro Viccini.
• Application of SUSY corrections (via Δ_b) to the charged Higgs production cross section improved. Now they are only applied to the bottom Yukawa coupling.
• Interpolation for Γ(H⁺ → tb) corrected. Thanks go to Marina Billoni.
• Bug in the initilization of the CKM matrix removed. Thanks go to Diego Guadagnoli.
• A warning is printed if the (g-2)_μ is unreliable due to a numerical instability. This can happen if two masses involved become degenerate.
• Including "FHRecord" simplified/improved.

• Bug in the NMFV evaluations (Higgs sector and BR(b → s γ)) removed (was only in version FH2.6.3).
• Bug in h → gg removed.

• Improved calculation of gg → φ, φ = h, H, A: more precise inclusion of interference of bottom loops with other contributions at large tanβ. Thanks go to Michael Spira.
• Augmented the Higgs-mass determination algorithm by a fully complex treatment of p².
  This could lead to (normally very) small differences in the Higgs boson masses.
  Correspondingly improved treatment of ZHiggs. Parameter points with extreme mixing of H and A are more stable now.
• The running b quark mass is now the DRbar mass (was the MSbar mass before). The shift in the mass should be below 5%.
• Improved calculation for BR(b → s γ): Resummation effects for large tanβ included.
• Interpolation of decay widths going from off- to on-shell decays.
• The parameters defining the CKM matrix are now the Wolfenstein parameters.
• Extended debug output (for FHDEBUG = 2).
• Some problems with the DRbar-to-OS transition for complex parameters removed.
• Configure adapted to g95 and gfortran.
• Bug in the call for NMFV parameters removed (was effective only from FH260 on).

• The invocation of the SM parameter settings has been simplified: A new, optional, call to set the SM parameters has been introduced, FHSetSMPara.
  All SM parameters, except m_t, but including the CKM parameters, are defined by this call. If this call is not made, default values are used, see the corresponding man page.
  The call to FHSetPara. has been shortened accordingly.
  The routine FHSetCKM does not exist anymore.
• Four new functions to retrieve data have been included:
  FHRetrieveFlags, FHRetrieveSMPara, FHRetrievePara, FHRetrieveNMFV.
  The new functions allow to retrieve "input data" from FeynHiggs. This is useful if the parameters are set via SLHA or benchmark planes.
• The definition of the Z-matrix has been changed slightly (see hep-ph/0611326, eq. (88) and before):
  Now a complex mass square is used as momentum; the Re[..] in the defintion of the diagonal entries has been dropped.
  Thanks go to Federico von der Pahlen and Karina Williams.
• The bottom quark mass input is now defined as m_b(m_b), default 4.20 GeV.
  The charm quark mass input is defined as m_c(m_c), default 1.286 GeV.
  The strange quark mass input is defined as m_s(2 GeV), default is 0.095 GeV.
• Running m_b(Q) (and other quark masses): bug removed, effects visible for Q > m_t.
  This also affects the Higgs production cross section bb → φ, φ = H, A with M_H,A > m_t.
  Thanks go to Robert Harlander, Wolfgang Mader, Michael Spira and Markus Warsinsky.
• α_s(Q): improved running for Higgs decays. Now we use everywhere the official PDG formula, see here.
• For the O(α_tα_s) corrections in the cMSSM: numerical instability removed for A_t → 0.
  Another numerical instability removed for X_t → 0.
• BR(t → H^± b): bug removed.
• The gluino mass, MGl, is now also an output in FHGetPara.

• The two-loop O(α_tα_s) corrections entering the Higgs-boson masses and mixings are now evaluated with the full complex phase dependence, see arXiv:0705.0746 [hep-ph].
• For the evaluation of the two-loop corrections in the precsence of complex phases a new 9th flag is introduced, see the FHSetFlags man pages. The options are:
  0: all corrections (asat, asab, atat, atab) are computed in the rMSSM
  1: only the cMSSM asat corrections are used
  2: the cMSSM asat corrections are combined with the remaining corrections in the rMSSM
  3: the cMSSM asat corrections are combined with the remaining corrections, interpolated in the complex phases
  The new default is 3.
• Self-defined parameter planes can be processed. The data is available in a two-dimensional grid format. An interpolation is used to obtain parameter values for any input. See the new man pages for the FHRecord (last line of man pages above).
• Predefined NUHM benchmark planes (in agreement with Cold Dark Matter measurements) are available, see arXiv:0709.0098 [hep-ph].
  The planes are available here.
• The production cross section gb → tH^- at the LHC is included. The cross section without Delta_b corrections are taken from hep-ph/0206121 (by T. Plehn) and hep-ph/0312286 (by E. Berger, T. Han, J. Jiang and T. Plehn).
  The cross sections are supplemented with the Delta_b correction factors.
• The BR(t → H⁺ b) is evaluated (and can be used for the production cross section for low charged Higgs boson masses).
• The BR(H⁺ → t^*b) is evaluated for off-shell top quarks.
• A bug in the computation of m_b(Q) for Q > m_t has been removed (thanks go to Karina Williams). This can lead to a noticible change in Gamma(H,A → bb) for M_H,A > m_t. The cross section evaluation with the radiation off bottom quarks, however, was ok.
• FeynHiggs now compiles under Windows (Cygwin), including the MathLink executable MFeynHiggs.
• FHGetPara now also includes the gluino mass (necessary e.g. if the new parameter planes are used).

• Inclusion of two new values for p2approx.
  p2approx = 4 is appropriate for large values of the heavy Higgs boson masses to ensure decoupling of effective couplings.
• Linking of FeynHiggs to C++ codes improved; example file (demo.cc) included.
• Routine for root finding improved to prevent infinite loops.
• configure improved (out-of-source-tree build possible).
• Bug in the FHUCC removed that prevented setting of M₁.

• The imaginary parts in the Higgs boson self-energies (at the one-loop) level are consistently taken into account for the evaluation of the masses and the Z-factors. Details can be found in hep-ph/0611326.
• The Z-factors are used to evaluate all Higgs decays and production cross sections (this was missing in a very few cases so far).
• The U-Matrix is now (again) evaluated from the 3x3 mass matrix evaluated at p²=m²_tree. By construction this matrix is unitary.
• The Higgs production cross section in association with bottom quarks extended to Higgs masses up to 1 TeV.
• The Higgs production cross section in association with the light stop has been included. The kinematics have been provided by Sherpa (thanks go to Frank Krauss).
• Formula for the EDM of the neutron improved.
• demo.m updated to match FeynHiggs2.5.
• Configuration extended and compilation tested for Mac OS X under g77 and xlf.

• Problem with diagonalization in the complex case solved (by implementing a completely new diagonalization routine).
• The Higgs masses are evaluated with two independent methods to detect potential numerical problems and possible instabilities. This can be especially relevant if masses are close to each other in the complex case.
• Decays of the neutral Higgs bosons are evaluated with the appropriate finite Z-factors (ZHiggs), see hep-ph/0003022 for the real case and hep-ph/0611326 for the complex case. The mixing matrix (UHiggs) is used to evaluate the effective couplings that can be used for the rotation of internally appearing Higgs bosons.
• Bug in DRbar to on-shell transition fixed. Now also the proper m_b value is used, rendering the transition more stable.
• demo.F updated to match FeynHiggs2.4.

• The calls to FHSetPara, FHHiggsCorr, FHConstraints have been modified slightly, see the man pages for details.
  You have to adjust these calls.
• Neutral Higgs boson decay are now evaluated with the full rotation to on-shell Higgs bosons, i.e. beyond the effective coupling approximation. The information is given as the new output matrix ZHiggs. (For the rotation of internal Higgs bosons the matrix UHiggs is provided as before.)
• The charged Higgs decay to light quarks is evaluated using running quark masses.
• The one-loop corrections to Higgs masses and mixing angles are now evaluated with the full 6x6 NMFV contributions (including off-diagonal mass terms and A terms).
  Corresponding new man pages are provided.
• The CKM matrix corrections are included.
  Corresponding new man pages are provided.
• The corresponding information can be passed with the new version of the SUSY Les Houches Accord 2, see also hep-ph/0605049 (by T. Hahn).
• The BR(b → s gamma) is evaluated, including NMFV effects, , see hep-ph/0512315 (by T. Hahn, W. Hollik, J. Illana, S. Penaranda) for details.
• As additional constraints M_W and sin²(theta_eff) are evaluated, including NMFV effects (full SM + Delta rho SUSY corrections).
  Note: Works correctly so far only for M_A > 150 GeV.

• FHCouplings is now fully addapted to handle complex parameters. Especially ttHiggs production cross sections are improved (thanks go to Federico v. d. Pahlen)
• slight (preliminary) change in the inclusion of the complex phases into the two-loop contribution. (Next update will include the full complex phase dependence for the strong two-loop corrections.)
• Phase conventions for UHiggs are chosen such that the results give a smooth limit to the real case (nothing wrong before, just a matter of convention).
• on-shell mass (either M_A or M_H+-) remains unchanged as it should (problem occured only for certain approximation calculations).

• The complex contributions to the Higgs boson self-energies at the one-loop level are taken into account. The Higgs boson pole masses are derived as the real parts of the complex poles of the complex propagator matrix.
• Correspondingly, the mixing matrix is now also complex. This is also taken into account in the Higgs boson couplings and decay widths.
• Negative entries are allowed for the squares of the soft SUSY-breaking parameters (for the diagonal entries for the sfermion mass matrices). The input is given as a negative mass, -m, that is then internally converted to -(m**2).
• slight (preliminary) change in the inclusion of the complex phases into the two-loop contribution.
• The 2-loop corrections to (g-2)_mu has been extended. All contributions described in hep-ph/0312264 (by S. Heinemeyer, D. Stöckinger, G. Weiglein) are now included by choosing the configure option --enable-full-g-2. If this option is not chosen only the leading two-loop are included.
• If the full version of (g-2)_mu is chosen, full precision is only guaranteed if quadruple precision is enabled (for the Intel Fortran compiler this is done with the configure option --quad.) The code might become slower...
• The call to FHCouplings has been modified slightly. It is extended by one variable (..., excl). See the man pages for details. If you call FHCouplings from your code you have to adjust this call.
• A bug in the coupling of the A to SM fermions and neutralinos has been corrected (thanks go to Federico von der Pahlen).
• Several small Fortran bugs have been removed.