Page Menu
Home
HEPForge
Search
Configure Global Search
Log In
Files
F7879223
No One
Temporary
Actions
View File
Edit File
Delete File
View Transforms
Subscribe
Mute Notifications
Award Token
Flag For Later
Size
22 KB
Subscribers
None
View Options
diff --git a/FixedOrderGen/src/PhaseSpacePoint.cc b/FixedOrderGen/src/PhaseSpacePoint.cc
index 41b21f1..032c139 100644
--- a/FixedOrderGen/src/PhaseSpacePoint.cc
+++ b/FixedOrderGen/src/PhaseSpacePoint.cc
@@ -1,657 +1,656 @@
#include "PhaseSpacePoint.hh"
#include <algorithm>
#include "HEJ/Constants.hh"
#include "HEJ/exceptions.hh"
#include "HEJ/kinematics.hh"
#include "HEJ/Particle.hh"
#include "HEJ/utility.hh"
#include "Process.hh"
#include "Subleading.hh"
using namespace HEJ;
namespace HEJFOG{
static_assert(
std::numeric_limits<double>::has_quiet_NaN,
"no quiet NaN for double"
);
constexpr double NaN = std::numeric_limits<double>::quiet_NaN();
HEJ::Event::EventData to_EventData(PhaseSpacePoint const & psp){
HEJ::Event::EventData result;
result.incoming = psp.incoming();
assert(result.incoming.size() == 2);
result.outgoing=psp.outgoing();
// technically Event::EventData doesn't have to be sorted,
// but PhaseSpacePoint should be anyway
assert(
std::is_sorted(
begin(result.outgoing), end(result.outgoing),
HEJ::rapidity_less{}
)
);
assert(result.outgoing.size() >= 2);
result.decays=psp.decays();
result.parameters.central= {NaN, NaN, psp.weight() };
return result;
}
namespace{
bool can_swap_to_uno(
HEJ::Particle const & p1, HEJ::Particle const & p2
){
return is_parton(p1)
&& p1.type != pid::gluon
&& p2.type == pid::gluon;
}
size_t count_gluons(std::vector<Particle>::const_iterator first,
std::vector<Particle>::const_iterator last){
return std::count_if(first, last, [](Particle const & p)
{return p.type == pid::gluon;});
}
/** assumes FKL configurations between first and last,
* else there can be a quark in a non-extreme position
* e.g. uno configuration gqg would pass
*/
bool can_change_to_qqx(
std::vector<Particle>::const_iterator first,
std::vector<Particle>::const_iterator last){
return 1 < count_gluons(first,last);
}
bool is_AWZ_proccess(Process const & proc){
return proc.boson && is_AWZ_boson(*proc.boson);
}
bool is_up_type(Particle const & part){
return HEJ::is_anyquark(part) && !(abs(part.type)%2);
}
bool is_down_type(Particle const & part){
return HEJ::is_anyquark(part) && abs(part.type)%2;
}
- /// true iff parton can couple to a W
- bool can_couple_to_W(Particle const & part, int const sign_W){
+ bool can_couple_to_W(Particle const & part, pid::ParticleID const W_id){
+ const int W_charge = W_id>0?1:-1;
return abs(part.type)<5
- && ( (sign_W*part.type > 0 && is_up_type(part))
- || (sign_W*part.type < 0 && is_down_type(part)) );
+ && ( (W_charge*part.type > 0 && is_up_type(part))
+ || (W_charge*part.type < 0 && is_down_type(part)) );
}
}
void PhaseSpacePoint::maybe_turn_to_subl(
double chance,
unsigned int const channels,
Process const & proc,
HEJ::RNG & ran
){
if(proc.njets <= 2) return;
assert(outgoing_.size() >= 2);
// decide what kind of subleading process is allowed
bool allow_uno = false;
bool allow_strange = true;
const size_t nout = outgoing_.size();
const bool can_be_uno_backward = (channels&Subleading::uno)
&& can_swap_to_uno(outgoing_[0], outgoing_[1]);
const bool can_be_uno_forward = (channels&Subleading::uno)
&& can_swap_to_uno(outgoing_[nout-1], outgoing_[nout-2]);
allow_uno = can_be_uno_backward || can_be_uno_forward;
bool allow_qqx = false;
if(is_AWZ_proccess(proc)) {
allow_qqx = (channels&Subleading::qqx)
&& can_change_to_qqx(outgoing_.cbegin(), outgoing_.cend());
- const int sign_W = *proc.boson>0?1:-1;
if(std::none_of(outgoing_.cbegin(), outgoing_.cend(),
- [sign_W](Particle const & p){ return can_couple_to_W(p, sign_W);})) {
+ [&proc](Particle const & p){ return can_couple_to_W(p, *proc.boson);})) {
// enforce qqx if A/W/Z can't couple somewhere else
assert(allow_qqx);
allow_uno = false;
chance = 1.;
// strange not allowed for W
if(abs(*proc.boson)== pid::Wp) allow_strange = false;
}
}
if(!allow_uno && !allow_qqx) return;
if(ran.flat() < chance){
weight_ /= chance;
if(allow_uno && !allow_qqx){
turn_to_uno(can_be_uno_backward, can_be_uno_forward, ran);
} else if (!allow_uno && allow_qqx) {
turn_to_qqx(allow_strange, ran);
} else {
assert( allow_uno && allow_qqx);
if(ran.flat() < 0.5) turn_to_uno(can_be_uno_backward, can_be_uno_forward, ran);
else turn_to_qqx(allow_strange, ran);
weight_ *= 2.;
}
} else weight_ /= 1 - chance;
}
void PhaseSpacePoint::turn_to_uno(
const bool can_be_uno_backward, const bool can_be_uno_forward,
HEJ::RNG & ran
){
if(!can_be_uno_backward && !can_be_uno_forward) return;
const size_t nout = outgoing_.size();
if(can_be_uno_backward && can_be_uno_forward){
if(ran.flat() < 0.5){
std::swap(outgoing_[0].type, outgoing_[1].type);
} else {
std::swap(outgoing_[nout-1].type, outgoing_[nout-2].type);
}
weight_ *= 2.;
} else if(can_be_uno_backward){
std::swap(outgoing_[0].type, outgoing_[1].type);
} else {
assert(can_be_uno_forward);
std::swap(outgoing_[nout-1].type, outgoing_[nout-2].type);
}
}
void PhaseSpacePoint::turn_to_qqx(const bool allow_stange, HEJ::RNG & ran){
/// find first and last gluon in FKL chain
auto first = std::find_if(outgoing_.begin(), outgoing_.end(),
[](Particle const & p){return p.type == pid::gluon;});
std::vector<Particle*> FKL_gluons;
for(auto p = first; p<outgoing_.end(); ++p){
if((*p).type == pid::gluon) FKL_gluons.push_back(&*p);
else if(is_quark(*p) || is_antiquark(*p)) break;
}
const size_t ng = FKL_gluons.size();
if(ng < 2)
throw std::logic_error("not enough gluons to create qqx");
// select flavour of quark
const double r1 = 2.*ran.flat()-1.;
const double max_flavour = allow_stange?n_f:n_f-1;
weight_ *= max_flavour*2;
int flavour = pid::down;
for (double sum = 1./max_flavour; sum < std::abs(r1); sum += 1./max_flavour)
++flavour;
flavour*=r1<0.?-1:1;
// select gluon for switch
const size_t idx = floor((ng-1) * ran.flat());
weight_ *= (ng-1);
FKL_gluons[idx]->type = ParticleID(flavour);
FKL_gluons[idx+1]->type = ParticleID(-flavour);
}
template<class ParticleMomenta>
fastjet::PseudoJet PhaseSpacePoint::gen_last_momentum(
ParticleMomenta const & other_momenta,
const double mass_square, const double y
) const {
std::array<double,2> pt{0.,0.};
for (auto const & p: other_momenta) {
pt[0]-= p.px();
pt[1]-= p.py();
}
const double mperp = sqrt(pt[0]*pt[0]+pt[1]*pt[1]+mass_square);
const double pz=mperp*sinh(y);
const double E=mperp*cosh(y);
return {pt[0], pt[1], pz, E};
}
PhaseSpacePoint::PhaseSpacePoint(
Process const & proc,
JetParameters const & jet_param,
HEJ::PDF & pdf, double E_beam,
double const subl_chance,
unsigned int const subl_channels,
ParticlesPropMap const & particles_properties,
HEJ::RNG & ran
)
{
assert(proc.njets >= 2);
if(proc.boson
&& particles_properties.find(*(proc.boson))
== particles_properties.end())
throw HEJ::missing_option("Boson "
+std::to_string(*(proc.boson))+" can't be generated: missing properties");
status_ = good;
weight_ = 1;
const int nout = proc.njets + (proc.boson?1:0);
outgoing_.reserve(nout);
// generate parton momenta
const bool is_pure_jets = !proc.boson;
auto partons = gen_LO_partons(
proc.njets, is_pure_jets, jet_param, E_beam, ran
);
// pre fill flavour with gluons
for(auto&& p_out: partons) {
outgoing_.emplace_back(Particle{pid::gluon, std::move(p_out), {}});
}
if(status_ != good) return;
// create boson
if(proc.boson){
const auto & boson_prop = particles_properties.at(*proc.boson);
auto boson(gen_boson(*proc.boson, boson_prop.mass, boson_prop.width, ran));
const auto pos = std::upper_bound(
begin(outgoing_),end(outgoing_),boson,rapidity_less{}
);
outgoing_.insert(pos, std::move(boson));
if(! boson_prop.decays.empty()){
const size_t boson_idx = std::distance(begin(outgoing_), pos);
decays_.emplace(
boson_idx,
decay_boson(outgoing_[boson_idx], boson_prop.decays, ran)
);
}
}
// normalisation of momentum-conserving delta function
weight_ *= pow(2*M_PI, 4);
/** @TODO
* uf (jet_param.min_pt) doesn't correspond to our final scale choice.
* The HEJ scale generators currently expect a full event as input,
* so fixing this is not completely trivial
*/
reconstruct_incoming(proc, subl_channels, pdf, E_beam, jet_param.min_pt, ran);
if(status_ != good) return;
// set outgoing states
most_backward_FKL(outgoing_).type = incoming_[0].type;
most_forward_FKL(outgoing_).type = incoming_[1].type;
maybe_turn_to_subl(subl_chance, subl_channels, proc, ran);
if(proc.boson) couple_boson(*proc.boson, ran);
}
double PhaseSpacePoint::gen_hard_pt(
int np , double ptmin, double ptmax, double y,
HEJ::RNG & ran
) {
// heuristic parameters for pt sampling
const double ptpar = ptmin + np/5.;
const double arg_small_y = atan((ptmax - ptmin)/ptpar);
const double y_cut = 3.;
const double r1 = ran.flat();
if(y < y_cut){
const double pt = ptmin + ptpar*tan(r1*arg_small_y);
const double temp = cos(r1*arg_small_y);
weight_ *= pt*ptpar*arg_small_y/(temp*temp);
return pt;
}
const double ptpar2 = ptpar/(1 + 5*(y-y_cut));
const double temp = 1. - std::exp((ptmin-ptmax)/ptpar2);
const double pt = ptmin - ptpar2*std::log(1-r1*temp);
weight_ *= pt*ptpar2*temp/(1-r1*temp);
return pt;
}
double PhaseSpacePoint::gen_soft_pt(int np, double max_pt, HEJ::RNG & ran) {
constexpr double ptpar = 4.;
const double r = ran.flat();
const double pt = max_pt + ptpar/np*std::log(r);
weight_ *= pt*ptpar/(np*r);
return pt;
}
double PhaseSpacePoint::gen_parton_pt(
int count, JetParameters const & jet_param, double max_pt, double y,
HEJ::RNG & ran
) {
constexpr double p_small_pt = 0.02;
if(! jet_param.peak_pt) {
return gen_hard_pt(count, jet_param.min_pt, max_pt, y, ran);
}
const double r = ran.flat();
if(r > p_small_pt) {
weight_ /= 1. - p_small_pt;
return gen_hard_pt(count, *jet_param.peak_pt, max_pt, y, ran);
}
weight_ /= p_small_pt;
const double pt = gen_soft_pt(count, *jet_param.peak_pt, ran);
if(pt < jet_param.min_pt) {
weight_=0.0;
status_ = not_enough_jets;
return jet_param.min_pt;
}
return pt;
}
std::vector<fastjet::PseudoJet> PhaseSpacePoint::gen_LO_partons(
int np, bool is_pure_jets,
JetParameters const & jet_param,
double max_pt,
HEJ::RNG & ran
){
if (np<2) throw std::invalid_argument{"Not enough partons in gen_LO_partons"};
weight_ /= pow(16.*pow(M_PI,3),np);
weight_ /= std::tgamma(np+1); //remove rapidity ordering
std::vector<fastjet::PseudoJet> partons;
partons.reserve(np);
for(int i = 0; i < np; ++i){
const double y = -jet_param.max_y + 2*jet_param.max_y*ran.flat();
weight_ *= 2*jet_param.max_y;
const bool is_last_parton = i+1 == np;
if(is_pure_jets && is_last_parton) {
constexpr double parton_mass_sq = 0.;
partons.emplace_back(gen_last_momentum(partons, parton_mass_sq, y));
break;
}
const double phi = 2*M_PI*ran.flat();
weight_ *= 2.0*M_PI;
const double pt = gen_parton_pt(np, jet_param, max_pt, y, ran);
if(weight_ == 0.0) return {};
partons.emplace_back(fastjet::PtYPhiM(pt, y, phi));
assert(jet_param.min_pt <= partons[i].pt());
assert(partons[i].pt() <= max_pt+1e-5);
}
// Need to check that at LO, the number of jets = number of partons;
fastjet::ClusterSequence cs(partons, jet_param.def);
auto cluster_jets=cs.inclusive_jets(jet_param.min_pt);
if (cluster_jets.size()!=unsigned(np)){
weight_=0.0;
status_ = not_enough_jets;
return {};
}
std::sort(begin(partons), end(partons), rapidity_less{});
return partons;
}
Particle PhaseSpacePoint::gen_boson(
HEJ::ParticleID bosonid, double mass, double width,
HEJ::RNG & ran
){
// Usual phase space measure
weight_ /= 16.*pow(M_PI, 3);
// Generate a y Gaussian distributed around 0
/// @TODO: magic number only for Higgs
/// @TODO better sampling for W
const double y = random_normal(1.6, ran);
const double r1 = ran.flat();
const double sH = mass*(
mass + width*tan(M_PI/2.*r1 + (r1-1.)*atan(mass/width))
);
auto p = gen_last_momentum(outgoing_, sH, y);
return Particle{bosonid, std::move(p), {}};
}
Particle const & PhaseSpacePoint::most_backward_FKL(
std::vector<Particle> const & partons
) const{
if(!HEJ::is_parton(partons[0])) return partons[1];
return partons[0];
}
Particle const & PhaseSpacePoint::most_forward_FKL(
std::vector<Particle> const & partons
) const{
const size_t last_idx = partons.size() - 1;
if(!HEJ::is_parton(partons[last_idx])) return partons[last_idx-1];
return partons[last_idx];
}
Particle & PhaseSpacePoint::most_backward_FKL(
std::vector<Particle> & partons
) const{
if(!HEJ::is_parton(partons[0])) return partons[1];
return partons[0];
}
Particle & PhaseSpacePoint::most_forward_FKL(
std::vector<Particle> & partons
) const{
const size_t last_idx = partons.size() - 1;
if(!HEJ::is_parton(partons[last_idx])) return partons[last_idx-1];
return partons[last_idx];
}
namespace {
/// partons are ordered: even = anti, 0 = gluon
ParticleID index_to_pid(size_t i){
if(!i) return pid::gluon;
return static_cast<ParticleID>(i%2?(i+1)/2:-i/2);
}
/// partons are ordered: even = anti, 0 = gluon
size_t pid_to_index(ParticleID id){
if(id==pid::gluon) return 0;
return id>0?id*2-1:abs(id)*2;
}
std::bitset<11> init_allowed(ParticleID const id){
if(abs(id) == pid::proton)
return ~0;
std::bitset<11> out = 0;
if(is_parton(id))
out[pid_to_index(id)] = 1;
return out;
}
/// decides which "index" (see index_to_pid) are allowed for process
std::bitset<11> allowed_quarks(ParticleID const boson){
std::bitset<11> allowed = ~0;
if(abs(boson) == pid::Wp){
// special case W:
// Wp: anti-down or up-type quark, no b/t -> 0001100110(1) = 205
// Wm: down or anti-up-type quark, no b/t -> 0010011001(1) = 307
allowed = boson>0?205:307;
}
return allowed;
}
}
/**
* checks which partons are allowed as initial state:
* 1. only allow what is given in the Runcard (p -> all)
* 2. A/W/Z require something to couple to
* a) no qqx => no incoming gluon
* b) 2j => no incoming gluon
* c) 3j => can couple OR is gluon => 2 gluons become qqx later
*/
std::array<std::bitset<11>,2> PhaseSpacePoint::filter_partons(
Process const & proc, unsigned int const subl_channels, HEJ::RNG & ran
){
std::array<std::bitset<11>,2> allowed_partons{
init_allowed(proc.incoming[0]),
init_allowed(proc.incoming[1])
};
bool const allow_qqx = subl_channels&Subleading::qqx;
// special case A/W/Z
if(is_AWZ_proccess(proc) && ((proc.njets < 4) || !allow_qqx)){
// all possible incoming states
auto allowed(allowed_quarks(*proc.boson));
if(proc.njets == 2 || !allow_qqx) allowed[0]=0;
// possible states per leg
std::array<std::bitset<11>,2> const maybe_partons{
allowed_partons[0]&allowed, allowed_partons[1]&allowed};
if(maybe_partons[0].any() && maybe_partons[1].any()){
// two options to get allowed initial state => choose one at random
const size_t idx = ran.flat() < 0.5;
allowed_partons[idx] = maybe_partons[idx];
// else choose the possible
} else if(maybe_partons[0].any()) {
allowed_partons[0] = maybe_partons[0];
} else if(maybe_partons[1].any()) {
allowed_partons[1] = maybe_partons[1];
} else{
throw std::invalid_argument{"Incoming state not allowed."};
}
}
return allowed_partons;
}
void PhaseSpacePoint::reconstruct_incoming(
Process const & proc, unsigned int const subl_channels,
HEJ::PDF & pdf, double E_beam,
double uf,
HEJ::RNG & ran
){
std::tie(incoming_[0].p, incoming_[1].p) = incoming_momenta(outgoing_);
// calculate xa, xb
const double sqrts=2*E_beam;
const double xa=(incoming_[0].p.e()-incoming_[0].p.pz())/sqrts;
const double xb=(incoming_[1].p.e()+incoming_[1].p.pz())/sqrts;
// abort if phase space point is outside of collider energy reach
if (xa>1. || xb>1.){
weight_=0;
status_ = too_much_energy;
return;
}
auto const & ids = proc.incoming;
std::array<std::bitset<11>,2> allowed_partons(
filter_partons(proc, subl_channels, ran));
for(size_t i = 0; i < 2; ++i){
if(ids[i] == pid::proton || ids[i] == pid::p_bar){
// pick ids according to pdfs
incoming_[i].type =
generate_incoming_id(i, i?xb:xa, uf, pdf, allowed_partons[i], ran);
} else {
assert(allowed_partons[i][pid_to_index(ids[i])]);
incoming_[i].type = ids[i];
}
}
assert(momentum_conserved(1e-7));
}
HEJ::ParticleID PhaseSpacePoint::generate_incoming_id(
size_t const beam_idx, double const x, double const uf,
HEJ::PDF & pdf, std::bitset<11> allowed_partons, HEJ::RNG & ran
){
std::array<double,11> pdf_wt;
pdf_wt[0] = allowed_partons[0]?fabs(pdf.pdfpt(beam_idx,x,uf,pid::gluon)):0.;
double pdftot = pdf_wt[0];
for(size_t i = 1; i < pdf_wt.size(); ++i){
pdf_wt[i] = allowed_partons[i]?4./9.*fabs(pdf.pdfpt(beam_idx,x,uf,index_to_pid(i))):0;
pdftot += pdf_wt[i];
}
const double r1 = pdftot * ran.flat();
double sum = 0;
for(size_t i=0; i < pdf_wt.size(); ++i){
if (r1 < (sum+=pdf_wt[i])){
weight_*= pdftot/pdf_wt[i];
return index_to_pid(i);
}
}
std::cerr << "Error in choosing incoming parton: "<<x<<" "<<uf<<" "
<<sum<<" "<<pdftot<<" "<<r1<<std::endl;
throw std::logic_error{"Failed to choose parton flavour"};
}
void PhaseSpacePoint::couple_boson(
HEJ::ParticleID const boson, HEJ::RNG & ran
){
if(abs(boson) != pid::Wp) return; // only matters for W
/// @TODO this could be use to sanity check gamma and Z
// find all possible quarks
- const int sign_W = boson>0?1:-1;
std::vector<Particle*> allowed_parts;
for(auto & part: outgoing_){
// Wp -> up OR anti-down, Wm -> anti-up OR down, no bottom
- if ( can_couple_to_W(part, sign_W) )
+ if ( can_couple_to_W(part, boson) )
allowed_parts.push_back(&part);
}
if(allowed_parts.size() == 0){
throw std::logic_error{"Found no parton for coupling with boson"};
}
// select one and flip it
size_t idx = 0;
if(allowed_parts.size() > 1){
/// @TODO more efficient sampling
/// old code: probability[i] = exp(parton[i].y - W.y)
idx = floor(ran.flat()*allowed_parts.size());
weight_ *= allowed_parts.size();
}
+ const int W_charge = boson>0?1:-1;
allowed_parts[idx]->type =
- static_cast<ParticleID>( allowed_parts[idx]->type - sign_W );
+ static_cast<ParticleID>( allowed_parts[idx]->type - W_charge );
}
double PhaseSpacePoint::random_normal(
double stddev,
HEJ::RNG & ran
){
const double r1 = ran.flat();
const double r2 = ran.flat();
const double lninvr1 = -log(r1);
const double result = stddev*sqrt(2.*lninvr1)*cos(2.*M_PI*r2);
weight_ *= exp(result*result/(2*stddev*stddev))*sqrt(2.*M_PI)*stddev;
return result;
}
bool PhaseSpacePoint::momentum_conserved(double ep) const{
fastjet::PseudoJet diff;
for(auto const & in: incoming()) diff += in.p;
for(auto const & out: outgoing()) diff -= out.p;
return nearby_ep(diff, fastjet::PseudoJet{}, ep);
}
Decay PhaseSpacePoint::select_decay_channel(
std::vector<Decay> const & decays,
HEJ::RNG & ran
){
double br_total = 0.;
for(auto const & decay: decays) br_total += decay.branching_ratio;
// adjust weight
// this is given by (channel branching ratio)/(chance to pick channel)
// where (chance to pick channel) =
// (channel branching ratio)/(total branching ratio)
weight_ *= br_total;
const double r1 = br_total*ran.flat();
double br_sum = 0.;
for(auto const & decay: decays){
br_sum += decay.branching_ratio;
if(r1 < br_sum) return decay;
}
throw std::logic_error{"unreachable"};
}
std::vector<Particle> PhaseSpacePoint::decay_boson(
HEJ::Particle const & parent,
std::vector<Decay> const & decays,
HEJ::RNG & ran
){
const auto channel = select_decay_channel(decays, ran);
if(channel.products.size() != 2){
throw HEJ::not_implemented{
"only decays into two particles are implemented"
};
}
std::vector<Particle> decay_products(channel.products.size());
for(size_t i = 0; i < channel.products.size(); ++i){
decay_products[i].type = channel.products[i];
}
// choose polar and azimuth angle in parent rest frame
const double E = parent.m()/2;
const double theta = 2.*M_PI*ran.flat();
const double cos_phi = 2.*ran.flat()-1.;
const double sin_phi = sqrt(1. - cos_phi*cos_phi); // Know 0 < phi < pi
const double px = E*cos(theta)*sin_phi;
const double py = E*sin(theta)*sin_phi;
const double pz = E*cos_phi;
decay_products[0].p.reset(px, py, pz, E);
decay_products[1].p.reset(-px, -py, -pz, E);
for(auto & particle: decay_products) particle.p.boost(parent.p);
return decay_products;
}
}
File Metadata
Details
Attached
Mime Type
text/x-diff
Expires
Tue, Nov 19, 7:44 PM (1 d, 7 h)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
3805900
Default Alt Text
(22 KB)
Attached To
rHEJ HEJ
Event Timeline
Log In to Comment