he problem A preview of GAMB Pat Scott mperial College London on behalf of the GAMB Collaboration http://gambit.hepforge.org Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? 3 things: Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? 3 things: A collaboration of about thirty theorists and experimentalists Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? 3 things: A collaboration of about thirty theorists and experimentalists A new public global fitting code Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? 3 things: A collaboration of about thirty theorists and experimentalists A new public global fitting code A program of physics analyses that we re carrying out using the code Pat Scott April 2015 F Madrid A preview of GAMB
he problem Let s begin with dessert. GAMB: he Global And Modular BSM nference ool So what is GAMB? 3 things: A collaboration of about thirty theorists and experimentalists A new public global fitting code A program of physics analyses that we re carrying out using the code First physics results and code release in a few months (i.e. late summer this year) Pat Scott April 2015 F Madrid A preview of GAMB
Outline he problem 1 he problem 2 3 Pat Scott April 2015 F Madrid A preview of GAMB
Outline he problem 1 he problem 2 3 Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches Question How do we know which models are in and which are out? Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches Question How do we know which models are in and which are out? Answer Combine the results from different searches Simplest method: take different exclusions, overplot them, conclude things are allowed or excluded Simplest BSM example: the scalar singlet model (Cline, Kainulainen, PS & Weniger, PRD, 1306.4710) log 10 λhs 0 1 2 3 XENON100 (2012) XENON100 5 XENON1 XENON100 20 Γh SS ΩS/ΩDM = 1 45 50 55 60 65 70 m S (GeV) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches hat s all well and good if there are only 2 parameters and few searches... Question What if there are many different constraints? Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches hat s all well and good if there are only 2 parameters and few searches... Question What if there are many different constraints? Answer Combine constraints in a statistically valid way composite likelihood log 10 λhs 0.5 0.0 0.5 1.0 1.5 Excluded by Γh SS Future 1σ CL ΩS/ΩDM = 1 Future 90% CL (Fermi + CA + Planck) Current 1σ CL (Fermi + WMAP7) (Cline, Kainulainen, PS & Weniger, PRD, 1306.4710) 2.0 2.0 2.5 3.0 3.5 log 10 (m S/GeV) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches hat s all well and good if there are only 2 parameters and few searches... Question What if there are many parameters? Pat Scott April 2015 F Madrid A preview of GAMB
he problem Combining searches hat s all well and good if there are only 2 parameters and few searches... Question What if there are many parameters? Answer Need to scan the parameter space (smart numerics) interpret the combined results (Bayesian / frequentist) project down to parameter planes of interest (marginalise / profile) global fits Pat Scott April 2015 F Madrid A preview of GAMB
he problem Beyond-the-Standard-Model Scanning Goals: 1 Given multiple theories, determine which fit the data better, and quantify how much better 2 Given a particular theory, determine which parameter combinations fit all experiments, and how well Pat Scott April 2015 F Madrid A preview of GAMB
he problem Beyond-the-Standard-Model Scanning Goals: 1 Given multiple theories, determine which fit the data better, and quantify how much better = model comparison 2 Given a particular theory, determine which parameter combinations fit all experiments, and how well = parameter estimation Pat Scott April 2015 F Madrid A preview of GAMB
he problem Beyond-the-Standard-Model Scanning Goals: 1 Given multiple theories, determine which fit the data better, and quantify how much better = model comparison 2 Given a particular theory, determine which parameter combinations fit all experiments, and how well = parameter estimation Why simple N/OU analyses are not enough... Only partial goodness of fit, no measure of convergence, no idea how to generalise to regions or whole space. Frequency/density of models in N/OU scans is not proportional to probability = no statistical meaning. statements about a theory s general ability to do one thing or another, based on such scans, are statistically invalid Pat Scott April 2015 F Madrid A preview of GAMB
Outline he problem 1 he problem 2 3 Pat Scott April 2015 F Madrid A preview of GAMB
he problem he LHC likelihood monster ALAS CMS ime per point: O(minute) in best cases GAMB Pat Scott April 2015 F Madrid A preview of GAMB
he problem he LHC likelihood monster ALAS CMS ime per point: O(minute) in best cases ime per point for global fits to converge: O(seconds) in worst cases GAMB Pat Scott April 2015 F Madrid A preview of GAMB
he problem he LHC likelihood monster ALAS CMS ime per point: O(minute) in best cases ime per point for global fits to converge: O(seconds) in worst cases GAMB Challenge: About 2 orders of magnitude too slow to actually include LHC data in global fits properly Pat Scott April 2015 F Madrid A preview of GAMB
he problem he LHC likelihood monster ALAS CMS ime per point: O(minute) in best cases ime per point for global fits to converge: O(seconds) in worst cases GAMB Challenge: About 2 orders of magnitude too slow to actually include LHC data in global fits properly More in Martin s presentation Pat Scott April 2015 F Madrid A preview of GAMB
he problem Doing genuinely model-independent DM pheno All experimental limits in terms of simplified models: effective WMP, one annihilation channel, etc = need something to apply limits to arbitrary DD couplings and D decay/annihilation branching fractions = must include accurate treatment of experimental effects Pat Scott April 2015 F Madrid A preview of GAMB
he problem Doing genuinely model-independent DM pheno All experimental limits in terms of simplified models: effective WMP, one annihilation channel, etc = need something to apply limits to arbitrary DD couplings and D decay/annihilation branching fractions = must include accurate treatment of experimental effects mpacts of new unstable particles (e.g. extra Higgs) are hard = need to simulate decays on the fly Pat Scott April 2015 F Madrid A preview of GAMB
he problem Doing genuinely model-independent DM pheno All experimental limits in terms of simplified models: effective WMP, one annihilation channel, etc = need something to apply limits to arbitrary DD couplings and D decay/annihilation branching fractions = must include accurate treatment of experimental effects mpacts of new unstable particles (e.g. extra Higgs) are hard = need to simulate decays on the fly Calculating relic densities for general models also challenging = want to feed in partial annihilation rates, co-annihilations, resonances, etc (not only set up model in LanHEP) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Doing genuinely model-independent DM pheno All experimental limits in terms of simplified models: effective WMP, one annihilation channel, etc = need something to apply limits to arbitrary DD couplings and D decay/annihilation branching fractions = must include accurate treatment of experimental effects mpacts of new unstable particles (e.g. extra Higgs) are hard = need to simulate decays on the fly Calculating relic densities for general models also challenging = want to feed in partial annihilation rates, co-annihilations, resonances, etc (not only set up model in LanHEP) nulike, gamlike, DDcalc, cascade sim Christoph s talk Pat Scott April 2015 F Madrid A preview of GAMB
he problem Parameter space heory space CMSSM, MSSM, Simplified Models BSM Want to do model comparison to actually work out which theory is the best... Challenge: How do easily adapt a global fit to different BSM theories? Pat Scott April 2015 F Madrid A preview of GAMB
he problem Parameter space heory space CMSSM, MSSM, Simplified Models BSM Want to do model comparison to actually work out which theory is the best... Challenge: How do easily adapt a global fit to different BSM theories? Somehow, we must recast things quickly to a new theory data likelihood functions scanning code housekeeping even predictions = a new, very abstract global fitting framework Pat Scott April 2015 F Madrid A preview of GAMB
Hitting the wall he problem ssues with current global fit codes: Strongly wedded to a few theories (e.g. constrained MSSM / msugra) Strongly wedded to a few theory calculators All datasets and observables basically hardcoded Rough or non-existent treatment of most experiments (astroparticle + collider especially) Sub-optimal statistical methods / search algorithms = already hitting the wall on theories, data & computational methods Pat Scott April 2015 F Madrid A preview of GAMB
Outline he problem 1 he problem 2 3 Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB: a second-generation global fit code GAMB: Global And Modular BSM nference ool Overriding principles of GAMB: flexibility and modularity General enough to allow fast definition of new datasets and theoretical models Plug and play scanning, physics and likelihood packages Extensive model database not just small modifications to constrained MSSM (NUHM, etc), and not just SUSY! Extensive observable/data libraries (likelihood modules) Many statistical options Bayesian/frequentist, likelihood definitions, scanning algorithms A smart and fast LHC likelihood calculator Massively parallel Full open-source code release Pat Scott April 2015 F Madrid A preview of GAMB
he problem he GAMB Collaboration 26 Members, 15 institutions, 9 countries 8 Experiments, 4 major theory codes Fermi-LA ALAS CA HESS LHCb cecube AMS-02 CDMS, DM-CE XENON/DARWN heory J. Conrad, J. Edsjö, G. Martinez P. Scott A. Buckley, P. Jackson, C. Rogan, A. Saavedra, M. White C. Balázs,. Bringmann, J. Conrad, M. White J. Conrad M. Chrza szcz, N. Serra J. Edsjö, C. Savage, P. Scott A. Putze L. Hsu J. Conrad P. Athron, C. Balázs,. Bringmann, J. Cornell, L. Dal, J. Edsjö, B. Farmer, A. Krislock, A. Kvellestad, M. Pato, F. Mahmoudi, A. Raklev, C. Savage, P. Scott, C. Weniger, M. White Pat Scott April 2015 F Madrid A preview of GAMB
he problem he GAMB Collaboration 26 Members, 15 institutions, 9 countries 8 Experiments, 4 major theory codes Fermi-LA ALAS CA HESS LHCb cecube AMS-02 CDMS, DM-CE XENON/DARWN heory J. Conrad, J. Edsjö, G. Martinez P. Scott A. Buckley, P. Jackson, C. Rogan, A. Saavedra, M. White C. Balázs,. Bringmann, J. Conrad, M. White J. Conrad M. Chrza szcz, N. Serra J. Edsjö, C. Savage, P. Scott A. Putze L. Hsu J. Conrad P. Athron, C. Balázs,. Bringmann, J. Cornell, L. Dal, J. Edsjö, B. Farmer, A. Krislock, A. Kvellestad, M. Pato, F. Mahmoudi, A. Raklev, C. Savage, P. Scott, C. Weniger, M. White Pat Scott April 2015 F Madrid A preview of GAMB
Modules he problem Physics Modules ColliderBit (Martin s talk) DarkBit (Christoph s talk) FlavBit flavour physics inc. g 2, b sγ, B decays (new channels, theory uncerts, LHCb likelihoods) SpecBit generic BSM spectrum object, providing RGE running, masses, mixings, etc via interchangeable interfaces to different RGE codes DecayBit decay widths for all relevant SM & BSM particles EWPOBit precision tests (mostly by interface to FeynHiggs, alt. SUSY-POPE) +ScannerBit: manages statistics, parameter sampling and optimisation algorithms Pat Scott April 2015 F Madrid A preview of GAMB
he problem Backends: mix and match GAMB modules consist of a number of standalone module functions Module functions can depend on each other, or they can require specific functions from backends Backends are external code libraries (DarkSUSY, FeynHiggs, etc) that include different functions GAMB automates and abstracts the interfaces to backends backend functions are tagged according to what they calculate with appropriate module design, different backends and their functions can be used interchangeably GAMB dynamically adapts to use whichever backends are actually present on a user s system (+ provides details of wtf it did of course) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Backends: mix and match GAMB modules consist of a number of standalone module functions Module functions can depend on each other, or they can require specific functions from backends Backends are external code libraries (DarkSUSY, FeynHiggs, etc) that include different functions GAMB automates and abstracts the interfaces to backends backend functions are tagged according to what they calculate with appropriate module design, different backends and their functions can be used interchangeably GAMB dynamically adapts to use whichever backends are actually present on a user s system (+ provides details of wtf it did of course) Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB: a toy example Pat Scott April 2015 F Madrid A preview of GAMB
he problem Dependency Resolution nuclear_params_fnq_parameters ype: ModelParameters Function: primary_parameters Module: nuclear_params_fnq DD_couplings ype: DarkBit::DD_couplings Function: DD_couplings_DarkSUSY sigma_sd_p ype: double Function: sigma_sd_p_simple capture_rate_sun ype: double DarkSUSY_5_1_1_init ype: void Function: DarkSUSY_5_1_1_init Module: BackendniBit sigma_s_p ype: double Function: sigma_s_p_simple Function: capture_rate_sun_constant_xsec equilibration_time_sun ype: double Function: equilibration_time_sun C79SL_data ype: nudata Function: C79SL_full annihilation_rate_sun ype: double Function: annihilation_rate_sun C79SL_loglike ype: double Function: C79SL_loglike DarkSUSY_Pointnit ype: bool Function: DarkSUSY_Pointnit_MSSM mwimp ype: double Function: mwimp_generic nulike_1_0_0_init ype: void Function: nulike_1_0_0_init Module: BackendniBit C79WH_data ype: nudata Function: C79WH_full C79WH_loglike ype: double Function: C79WH_loglike cecube_likelihood ype: double Function: C79_loglike MSSM25atQ_parameters ype: ModelParameters Function: primary_parameters MSSM_spectrum ype: easlha Module: MSSM25atQ Function: get_mssm_spectrum_from_file H_ProcessCatalog ype: DarkBit::H_ProcessCatalog Function: H_ProcessCatalog_CMSSM sigmav ype: double Function: sigmav_late_universe nuyield_ptr ype: nuyield_functype Function: nuyield_from_ds C79WL_data ype: nudata Function: C79WL_full C79WL_loglike ype: double Function: C79WL_loglike DarkMatter_D ype: DarkMatter_D_type Function: DarkMatter_D_MSSM25atQ pi_plus_decay_rates ype: Decayable::Entry Function: pi_plus_decays eta_decay_rates ype: Decayable::Entry Function: eta_decays rho_0_decay_rates ype: Decayable::Entry Function: rho_0_decays rho_minus_decay_rates ype: Decayable::Entry Function: rho_minus_decays rho_plus_decay_rates ype: Decayable::Entry Function: rho_plus_decays omega_decay_rates ype: Decayable::Entry Function: omega_decays gluino_decay_rates ype: Decayable::Entry Function: gluino_decays h0_2_decay_rates ype: Decayable::Entry Function: h0_2_decays A0_decay_rates ype: Decayable::Entry Function: A0_decays Hplus_decay_rates ype: Decayable::Entry Function: Hplus_decays stop_1_decay_rates ype: Decayable::Entry Function: stop_1_decays stop_2_decay_rates ype: Decayable::Entry Function: stop_2_decays sbottom_1_decay_rates ype: Decayable::Entry Function: sbottom_1_decays sbottom_2_decay_rates ype: Decayable::Entry Function: sbottom_2_decays MSSM78atQ_parameters ype: ModelParameters Function: MSSM78atQ_parameters Module: MSSM25atQ SUSY_H_1_4_init ype: void Function: SUSY_H_1_4_init Module: BackendniBit sdown_l_decay_rates ype: Decayable::Entry Function: sdown_l_decays sup_l_decay_rates ype: Decayable::Entry Function: sup_l_decays StandardModel_SLHA2_parameters ype: ModelParameters Function: primary_parameters Module: StandardModel_SLHA2 SMNPUS ype: SMnputs Function: get_smnpus Module: SpecBit MSSM_spectrum ype: const SMplusUV* Function: get_mssmatq_spectrum Module: SpecBit sup_r_decay_rates ype: Decayable::Entry Function: sup_r_decays decay_rates ype: Decayable Function: all_decays scharm_r_decay_rates ype: Decayable::Entry Function: scharm_r_decays sdown_r_decay_rates ype: Decayable::Entry Function: sdown_r_decays scharm_l_decay_rates ype: Decayable::Entry Function: scharm_l_decays sstrange_l_decay_rates ype: Decayable::Entry Function: sstrange_l_decays sstrange_r_decay_rates ype: Decayable::Entry Function: sstrange_r_decays Higgs_decay_rates ype: Decayable::Entry Function: MSSM_h0_1_decays W_minus_decay_rates ype: Decayable::Entry Function: W_minus_decays W_plus_decay_rates ype: Decayable::Entry Function: W_plus_decays Z_decay_rates ype: Decayable::Entry Function: Z_decays t_decay_rates ype: Decayable::Entry Function: t_decays tbar_decay_rates ype: Decayable::Entry Function: tbar_decays mu_minus_decay_rates ype: Decayable::Entry Function: mu_minus_decays mu_plus_decay_rates ype: Decayable::Entry Function: mu_plus_decays tau_minus_decay_rates ype: Decayable::Entry Function: tau_minus_decays tau_plus_decay_rates ype: Decayable::Entry Function: tau_plus_decays pi_0_decay_rates ype: Decayable::Entry Function: pi_0_decays pi_minus_decay_rates ype: Decayable::Entry Function: pi_minus_decays Module functions and backend functions get arranged into a dependency tree Starting with requested observables and likelihoods, fills each dependency and backend requirement Obeys rules at each step: allowed models, allowed backends, constraints from input file, etc tree constitutes a directed acyclic graph GAMB uses graph-theoretic methods to solve the graph to determine function evaluation order Pat Scott April 2015 F Madrid A preview of GAMB
he problem Dependency Resolution DarkSUSY_5_1_1_init MSSM_spectrum Module functions and backend ype: functions void ype: get easlha arranged into Function: DarkSUSY_5_1_1_init Function: get_mssm_spectrum_from_file Module: BackendniBit a dependency tree nuclear_params_fnq_parameters DarkSUSY_Pointnit DarkMatter_D ype: ModelParameters ype: bool ype: DarkMatter_D_type Function: primary_parameters Function: DarkSUSY_Pointnit_MSSM Function: DarkMatter_D_MSSM25atQ Module: nuclear_params_fnq Starting with requested observables and likelihoods, fills DD_couplings H_ProcessCatalog ype: DarkBit::DD_couplings ype: DarkBit::H_ProcessCatalog Function: DD_couplings_DarkSUSY Function: H_ProcessCatalog_CMSSM each dependency and backend requirement ype: double Function: mwimp_generic Obeys rules at each step: allowed models, allowed sigma_sd_p sigma_s_p sigmav ype: double ype: double ype: double backends, constraints Function: from sigma_sd_p_simple input Function: sigma_s_p_simple file, etcfunction: sigmav_late_universe capture_rate_sun ype: double tree constitutes a directed Function: capture_rate_sun_constant_xsec acyclic graph nuyield_ptr ype: nuyield_functype Function: nuyield_from_ds equilibration_time_sun ype: double GAMB uses graph-theoretic Function: equilibration_time_sun methods to solve the graph to determine functionannihilation_rate_sun evaluation nulike_1_0_0_init order ype: double Function: annihilation_rate_sun mwimp ype: void Function: nulike_1_0_0_init Module: BackendniBit MSSM25atQ_parameters ype: ModelParameters Function: primary_parameters Module: MSSM25atQ pi_plus_decay_rates ype: Decayable::Entry Function: pi_plus_decays eta_decay_ra ype: Decayable:: Function: eta_dec Module: DecayB C79SL_data ype: nudata Function: C79SL_full C79WH_data ype: nudata Function: C79WH_full C79WL_data ype: nudata Function: C79WL_full C79SL_loglike ype: double Function: C79SL_loglike C79WH_loglike ype: double Function: C79WH_loglike cecube_likelihood ype: double Function: C79_loglike C79WL_loglike ype: double Function: C79WL_loglike Pat Scott April 2015 F Madrid A preview of GAMB
he problem Hierarchical Model Database Models are defined by their parameters and relations to each other Models can inherit from parent models Points in child models can be automatically translated to ancestor models Friend models also allowed (cross-family translation) Model dependence of every module/backend function is tracked = maximum safety, maximum reuse MSSM78atQ MSSM25atQ SingletDM StandardModel_SLHA2 nuclear_params_fnq nuclear_params_sigma0_sigmapin nuclear_params_sigmas_sigmapin DMHalo_base_demo Gaussian_Halo_demo NormalDist SomeOther_Halo_demo CMSSM_demo test_parent_ MSSM_demo CMSSM demo WOHDM_demo_parent WOHDM_demo WOHDM_sub_demo MSSM78atMGU CMSSM extracmssm Pat Scott April 2015 F Madrid A preview of GAMB
he problem nterface: yaml file Basic interface for a scan is a YAML initialisation file - specify parameters, ranges, priors - select likelihood components - select other observables to calculate - define generic rules for how to fill dependencies - define generic rules for options to be passed to module functions - set global options (scanner, errors/warnings, logging behaviour, etc) Pat Scott April 2015 F Madrid A preview of GAMB
he problem nterface: yaml file Basic interface for a scan is a YAML initialisation file - specify parameters, ranges, priors - select likelihood components - select other observables to calculate - define generic rules for how to fill dependencies - define generic rules for options to be passed to module functions - set global options (scanner, errors/warnings, logging behaviour, etc) Pat Scott April 2015 F Madrid A preview of GAMB
he problem nterface: yaml file Basic interface for a scan is a YAML initialisation file - specify parameters, ranges, priors - select likelihood components - select other observables to calculate - define generic rules for how to fill dependencies - define generic rules for options to be passed to module functions - set global options (scanner, errors/warnings, logging behaviour, etc) Pat Scott April 2015 F Madrid A preview of GAMB
he problem nterface: yaml file Basic interface for a scan is a YAML initialisation file - specify parameters, ranges, priors - select likelihood components - select other observables to calculate - define generic rules for how to fill dependencies - define generic rules for options to be passed to module functions - set global options (scanner, errors/warnings, logging behaviour, etc) Pat Scott April 2015 F Madrid A preview of GAMB
he problem nterface: yaml file Basic interface for a scan is a YAML initialisation file - specify parameters, ranges, priors - select likelihood components - select other observables to calculate - define generic rules for how to fill dependencies - define generic rules for options to be passed to module functions - set global options (scanner, errors/warnings, logging behaviour, etc) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Expansion: adding new functions Adding a new module function is easy: 1 Declare the function to GAMB in a module s rollcall header Choose a capability Declare any dependencies Declare any backend requirements Declare any specific allowed models other more advanced declarations also available 2 Write the function as a simple C++ function (one argument: the result) Pat Scott April 2015 F Madrid A preview of GAMB
he problem Other nice technical features Scanners: MultiNest, Diver (diff. evolution), PKAA (genetic algorithms), GreA (MCMC) Statistics: Bayesian, Profile Likelihood, later full Neyman Mixed-mode MP + openmp, mostly automated diskless generalisation of various Les Houches Accords BOSS: dynamic loading of C++ classes from backends (!) all-in or module standalone modes easily implemented from single cmake script automatic getters for obtaining, configuring + compiling backends 1 flexible output streams (ASC, databases, binary,... ) more more more... 1 if a backend breaks, won t compile and/or kills your dog, blame the authors (not us... unless we are the authors... ) Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
he problem GAMB vs the rest in a nutshell Aspect GAMB MasterCode SuperBayeS Fittino Rizzo et al. Design Modular, Adaptive Monolithic Monolithic ( )Monolithic Monolithic Statistics Frequentist, Bayesian Frequentist Freq./Bayes. Frequentist None Scanners Differential evolution, genetic algorithms, random forests, t-walk, t- nest, particle swarm, nested sampling, MCMC, gradient descent heories (p)mssm-25, CMSSM±ɛ, GMSB, AMSB, gaugino mediation, E6MSSM, NMSSM, BMSSM, PQMSSM, effective operators, idm, XDM, ADM, UED, Higgs portals/extended Higgs sectors Astroparticle Event-level: cecube, Fermi, LUX, XENON, CDMS, DM-CE. AMS-02, COUPP, KMS, CRESS, CoGeN, SMPLE, PAMELA, Planck, HESS. Predictions: CA, DARWN, GAPS LHC ALAS+CMS multi-analysis with neural net and fast detector simulation. Higgs multi-channel with correlations and no SM assumptions. Full flavour inc. complete B X sll and B K ll angular set. SM, theory m t, m b, α s, α EM, DM halo, hadronic and related matrix elements, detector responses, uncerts. QCD+EW corrections (LHC+DM signal+bg), astro BGs, cosmic ray hadronisation, coalescence and p gation. Nested sampling, MCMC, grad. descent CMSSM±ɛ LUX, XENON ALAS resim, HiggsSignals, basic flavour. m t, m Z, α EM, hadronic matrix elements Nested sampling, MCMC (p)mssm-15, CMSSM±ɛ, mued Fermi, cecube, XENON ALAS direct sim, Higgs mass only, basic flavour. m t, m b, α s, α EM, DM halo, hadronic matrix elems. MCMC None (random) CMSSM±ɛ Fermi, HESS, XENON ALAS resim, HiggsSignals, basic flavour. m t (p)mssm-19 Event-level: Fermi. cecube, CA ALAS+CMS +evatron direct sim, basic flavour. None Pat Scott April 2015 F Madrid A preview of GAMB
Closing remarks he problem Robust analysis of dark matter and BSM physics requires multi-messenger global fits GAMB is coming: Global fits to many models for the first time Better global fits to familiar ones Highly modular, usable and extendable public code Faster, more complete and more consistent theory explorations + experimental analysis prototyping Pat Scott April 2015 F Madrid A preview of GAMB
he problem Pat Scott April 2015 F Madrid A preview of GAMB