jiri/LA_library
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

tensor: bugfixes and hermiticity enforcement

Browse Source
This commit is contained in:
Jiri Pittner
2025-11-19 14:58:45 +01:00
parent 78569ca701
commit c067029b1d
4 changed files with 199 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
la_traits.h
View File

@@ -306,6 +306,8 @@ static void deallocate(std::complex<C> &x) {};
static inline std::complex<C> conjugate(const std::complex<C> &x) {return std::complex<C>(x.real(),-x.imag());};
static inline C realpart(const std::complex<C> &x) {return x.real();}
static inline C imagpart(const std::complex<C> &x) {return x.imag();}
static inline void setrealpart(std::complex<C> &x, const C &y) {reinterpret_cast<C(&)[2]>(x)[0]=y;}
static inline void setimagpart(std::complex<C> &x, const C &y) {reinterpret_cast<C(&)[2]>(x)[1]=y;}
};
@@ -364,6 +366,8 @@ static void deallocate(C &x) {};
static inline C conjugate(const C &x) {return x;};
static inline C realpart(const C &x) {return x;}
static inline C imagpart(const C &x) {return 0;}
static inline void setrealpart(C &x, const C &y) {x=y;}
static inline void setimagpart(C &x, const C &y) {}
};

85
t.cc
View File

@@ -4263,7 +4263,7 @@ for(int k=0; k<nn; ++k) for(int l=0; l<nn; ++l) for(int m=0; m<nn; ++m) for(int
cout <<"Error = "<<(z-zz).norm()<<endl;
}
if(1)
if(0)
{
int nn;
cin>>nn;
@@ -4309,5 +4309,88 @@ cout <<"Error = "<<(z-zzz).norm()<<endl;
}
if(0)
{
//check symmetrizer/antisymmetrizer in general case
int r,n,sym;
cin>>r>>n>>sym;
NRVec<INDEXGROUP> shape(3);
shape[0].number=2;
shape[0].symmetry=0;
shape[0].range=n+1;
shape[0].offset=0;
shape[1].number=r;
shape[1].symmetry= sym;
shape[1].range=n;
shape[1].offset=0;
shape[2].number=2;
shape[2].symmetry=0;
shape[2].range=n+2;
shape[2].offset=0;
Tensor<complex<double> > x(shape); x.randomize(1.);
x.defaultnames();
cout <<"x= "<<x.shape << " "<<x.names<<endl;
Tensor<complex<double> > xf=x.flatten(1);
cout <<"xf= "<<xf.shape << " "<<xf.names<<endl;
Tensor<complex<double> > xxx=x.unwind_index_group(1);
cout <<"xxx= "<<xxx.shape<<" "<<xxx.names<<endl;
INDEXLIST il(r);
for(int i=0; i<r; ++i) il[i]= {1+i,0};
Tensor<complex<double> > xx = xf.merge_indices(il,sym);
cout <<"xx = "<<xx.shape<< " "<<xx.names<<endl;
cout <<"Error = "<<(xx-xxx).norm()<<endl;
}
if(1)
{
int nn;
cin>>nn;
int r=4;
NRVec<INDEXGROUP> s(r);
for(int i=0; i<r; ++i)
{
s[i].number=1;
s[i].symmetry=0;
s[i].range=nn;
s[i].offset=0;
}
Tensor<complex<double> > x(s); x.randomize(1.);
INDEXNAME xlist[] = {"i","j","k","l"};
x.names=NRVec<INDEXNAME>(xlist);
Tensor<complex<double> > y(s); y.randomize(1.);
INDEXNAME ylist[] = {"n","m","j","i"};
y.names=NRVec<INDEXNAME>(ylist);
Tensor<complex<double> >z(s); z.clear();
INDEXNAME zlist[] = {"k","l","m","n"};
z.names=NRVec<INDEXNAME>(zlist);
Tensor<complex<double> >zz(z);
z.add_permuted_contractions("k/l,m",x,y,1,0,false,false);
zz.copyonwrite();
for(int k=0; k<nn; ++k) for(int l=0; l<nn; ++l) for(int m=0; m<nn; ++m) for(int n=0; n<nn; ++n)
{
complex<double> s=0;
for(int i=0; i<nn; ++i) for(int j=0; j<nn; ++j) s += x(i,j,k,l)*y(n,m,j,i);
zz.lhs(k,l,m,n) = s;
}
Tensor<complex<double> > zzz(s);
for(int k=0; k<nn; ++k) for(int l=0; l<nn; ++l) for(int m=0; m<nn; ++m) for(int n=0; n<nn; ++n)
{
zzz.lhs(k,l,m,n) = (zz(k,l,m,n)-zz(l,k,m,n)+zz(m,k,l,n));
}
cout <<"Error = "<<(z-zzz).norm()<<endl;
}
}//main

92
tensor.cc
View File

@@ -24,6 +24,7 @@
#include <complex>
#include "bitvector.h"
#include <string.h>
#include <bits/stdc++.h>
namespace LA {
@@ -899,7 +900,7 @@ template<typename T>
static void flatten_callback(const SUPERINDEX &I, T *v)
{
FLATINDEX J = superindex2flat(I);
//std::cout <<"TEST flatten_callback: from "<<JP<<" TO "<<J<<std::endl;
//std::cout <<"TEST flatten_callback: "<<J<<std::endl;
*v = (*help_tt<T>)(J); //rhs operator() generates the redundant elements for the unwinded lhs tensor
}
//
@@ -1432,6 +1433,7 @@ template<typename T>
static const PermutationAlgebra<int,T> *help_pa;
static bool help_inverse;
static bool help_conjugate;
template<typename T>
static T help_alpha;
@@ -1443,7 +1445,10 @@ FLATINDEX J = superindex2flat(I);
for(int p=0; p<help_pa<T>->size(); ++p)
{
FLATINDEX Jp = J.permuted((*help_pa<T>)[p].perm,help_inverse);
*v += help_alpha<T> * (*help_pa<T>)[p].weight * (*help_tt<T>)(Jp);
bool conj=false;
if(help_conjugate) conj= ((*help_pa<T>)[p].perm).parity()<0;
T tmp= (*help_tt<T>)(Jp);
*v += help_alpha<T> * (*help_pa<T>)[p].weight * (conj ? LA_traits<T>::conjugate(tmp) : tmp);
}
}
@@ -1476,7 +1481,7 @@ for(int p=0; p<help_pa<T>->size(); ++p)
template<typename T>
void Tensor<T>::apply_permutation_algebra(const Tensor<T> &rhs, const PermutationAlgebra<int,T> &pa, bool inverse, T alpha, T beta)
void Tensor<T>::apply_permutation_algebra(const Tensor<T> &rhs, const PermutationAlgebra<int,T> &pa, bool inverse, T alpha, T beta, bool conjugate)
{
if(beta!=(T)0) {if(beta!=(T)1) *this *= beta;} else clear();
if(alpha==(T)0) return;
@@ -1500,6 +1505,7 @@ if(rhs.is_named())
help_tt<T> = &rhs;
help_pa<T> = &pa;
help_inverse = inverse;
help_conjugate= LA_traits<T>::is_complex() ? conjugate : false;
help_alpha<T> = alpha;
loopover(permutationalgebra_callback);
}
@@ -1882,6 +1888,7 @@ if(!basicperm.is_valid()) laerror("internal error in merge_indices");
//prepare permutation algebra
PermutationAlgebra<int,T> pa;
bool doconjugate=false;
if(sym==0)
{
pa.resize(1);
@@ -1890,6 +1897,7 @@ if(sym==0)
}
else
{
doconjugate = sym>=2||sym<= -2;
PermutationAlgebra<int,int> sa = sym>0 ? symmetrizer<int>(il.size()) : antisymmetrizer<int>(il.size());
//std::cout <<"SA = "<<sa<<std::endl;
pa.resize(sa.size());
@@ -1905,7 +1913,7 @@ else
//std::cout <<"Use PA = "<<pa<<std::endl;
Tensor<T> r(newshape);
r.apply_permutation_algebra(*this,pa,false,(T)1/(T)pa.size(),0);
r.apply_permutation_algebra(*this,pa,false,(T)1/(T)pa.size(),0,doconjugate);
if(is_named()) r.names=names.permuted(basicperm,true);
return r;
}
@@ -2200,6 +2208,82 @@ if(is_named())
}
static void checkindex(const NRVec<INDEXGROUP> &shape, const SUPERINDEX &I, bool &zeroreal, bool &zeroimag)
{
#ifdef DEBUG
if(shape.size()!=I.size()) laerror("inconsistent shape and index in checkindex");
#endif
zeroreal=zeroimag=false;
for(int g=0; g<I.size(); ++g)
{
#ifdef DEBUG
if(I[g].size()!=shape[g].number) laerror("inconsistent2 shape and index in checkindex");
#endif
if(shape[g].symmetry == 2 || shape[g].symmetry == -2)
{
bool sameindex=false;
for(int i=1; i<shape[g].number; ++i) for(int j=0; j<i; ++j)
if(I[g][i]==I[g][j])
{
sameindex=true;
goto checkfailed; //for this group, but do the remaining ones for further restrictions
}
checkfailed:
if(sameindex)
{
if(shape[g].symmetry == 2) zeroimag=true;
if(shape[g].symmetry == -2) zeroreal=true;
}
}
}
}
template<typename T>
static void hermiticity_callback(const SUPERINDEX &I, T *v)
{
bool zeroimag;
bool zeroreal;
checkindex(help_t<T>->shape,I,zeroreal,zeroimag);
if(zeroreal) LA_traits<T>::setrealpart(*v,0);
if(zeroimag) LA_traits<T>::setimagpart(*v,0);
}
template<typename T>
static void hermiticity_callback2(const SUPERINDEX &I, const T *v)
{
bool zeroimag;
bool zeroreal;
checkindex(help_tt<T>->shape,I,zeroreal,zeroimag);
if(zeroreal && LA_traits<T>::realpart(*v)!=(T)0 || zeroimag && LA_traits<T>::imagpart(*v)!=(T)0) throw true;
}
template<typename T>
void Tensor<T>::enforce_hermiticity()
{
if(!has_hermiticity()) return;
help_t<T> = this;
loopover(hermiticity_callback);
}
template<typename T>
bool Tensor<T>::fulfills_hermiticity() const
{
if(!has_hermiticity()) return true;
help_tt<T> = this;
try {
constloopover(hermiticity_callback2);
}
catch(bool failed) {return false;}
return true;
}
template class Tensor<double>;
template class Tensor<std::complex<double> >;

38
tensor.h
View File

@@ -50,7 +50,7 @@
namespace LA {
template<typename T>
inline T signeddata(const int sgn, const T data, const bool lhs=false)
static inline T signeddata(const int sgn, const T &data)
{
if(LA_traits<T>::is_complex()) //condition known at compile time
{
@@ -69,9 +69,6 @@ if(LA_traits<T>::is_complex()) //condition known at compile time
return -LA_traits<T>::conjugate(data);
break;
case 0:
#ifdef DEBUG
if(lhs) laerror("dereferencing lhs Signedpointer to nonexistent tensor element");
#endif
return 0;
break;
}
@@ -81,9 +78,6 @@ if(LA_traits<T>::is_complex()) //condition known at compile time
{
if(sgn>0) return data;
if(sgn<0) return -data;
#ifdef DEBUG
if(sgn==0 && lhs) laerror("dereferencing lhs Signedpointer to nonexistent tensor element");
#endif
return 0;
}
@@ -273,12 +267,24 @@ public:
void resize(const NRVec<INDEXGROUP> &s) {shape=s; data.resize(calcsize()); calcrank(); names.clear();};
void deallocate() {data.resize(0); shape.resize(0); groupsizes.resize(0); cumsizes.resize(0); names.resize(0);};
inline Signedpointer<T> lhs(const SUPERINDEX &I) {int sign; LA_largeindex i=index(&sign,I); return Signedpointer<T>(&data[i],sign);};
inline T operator()(const SUPERINDEX &I) const {int sign; LA_largeindex i=index(&sign,I); return signeddata(sign,data[i]);};
inline Signedpointer<T> lhs(const FLATINDEX &I) {int sign; LA_largeindex i=index(&sign,I); return Signedpointer<T>(&data[i],sign);};
inline T operator()(const FLATINDEX &I) const {int sign; LA_largeindex i=index(&sign,I); return signeddata(sign,data[i]);};
inline Signedpointer<T> lhs(LA_index i1...) {va_list args; int sign; LA_largeindex i; va_start(args,i1); i= vindex(&sign, i1,args); return Signedpointer<T>(&data[i],sign); };
inline T operator()(LA_index i1...) const {va_list args; ; int sign; LA_largeindex i; va_start(args,i1); i= vindex(&sign, i1,args); return signeddata(sign,data[i]);};
inline Signedpointer<T> lhs(const SUPERINDEX &I) {int sign; LA_largeindex i=index(&sign,I);
#ifdef DEBUG
if(sign==0) laerror("l-value pointer to nonexistent tensor element");
#endif
return Signedpointer<T>(&data[i],sign);};
inline T operator()(const SUPERINDEX &I) const {int sign; LA_largeindex i=index(&sign,I); if(sign==0) return 0; else return signeddata(sign,data[i]);};
inline Signedpointer<T> lhs(const FLATINDEX &I) {int sign; LA_largeindex i=index(&sign,I);
#ifdef DEBUG
if(sign==0) laerror("l-value pointer to nonexistent tensor element");
#endif
return Signedpointer<T>(&data[i],sign);};
inline T operator()(const FLATINDEX &I) const {int sign; LA_largeindex i=index(&sign,I); if(sign==0) return 0; else return signeddata(sign,data[i]);};
inline Signedpointer<T> lhs(LA_index i1...) {va_list args; int sign; LA_largeindex i; va_start(args,i1); i= vindex(&sign, i1,args);
#ifdef DEBUG
if(sign==0) laerror("l-value pointer to nonexistent tensor element");
#endif
return Signedpointer<T>(&data[i],sign); };
inline T operator()(LA_index i1...) const {va_list args; ; int sign; LA_largeindex i; va_start(args,i1); i= vindex(&sign, i1,args); if(sign==0) return 0; else return signeddata(sign,data[i]);};
inline Tensor& operator=(const Tensor &rhs) {myrank=rhs.myrank; shape=rhs.shape; groupsizes=rhs.groupsizes; cumsizes=rhs.cumsizes; data=rhs.data; names=rhs.names; return *this;};
@@ -332,7 +338,9 @@ public:
void put(int fd, bool with_names=false) const;
void get(int fd, bool with_names=false);
inline void randomize(const typename LA_traits<T>::normtype &x) {if(has_hermiticity()) laerror("randomization does not support correct treatment of hermitean/antihermitean index groups"); data.randomize(x);};
void enforce_hermiticity(); //zero out real/imag parts for repeated indices appropriately
bool fulfills_hermiticity() const; //check it is so
inline void randomize(const typename LA_traits<T>::normtype &x) {data.randomize(x); enforce_hermiticity();};
void loopover(void (*callback)(const SUPERINDEX &, T *)); //loop over all elements
void constloopover(void (*callback)(const SUPERINDEX &, const T *)) const; //loop over all elements
@@ -373,7 +381,7 @@ public:
Tensor innercontraction(const NRVec<INDEXNAME> &nl1, const NRVec<INDEXNAME> &nl2) const {return innercontraction(findindexlist(nl1),findindexlist(nl2));};
Tensor innercontraction(const INDEXNAME &n1, const INDEXNAME &n2) const {return innercontraction(findindex(n1),findindex(n2));};
void apply_permutation_algebra(const Tensor &rhs, const PermutationAlgebra<int,T> &pa, bool inverse=false, T alpha=1, T beta=0); //general (not optimally efficient) symmetrizers, antisymmetrizers etc. acting on the flattened index list:
void apply_permutation_algebra(const Tensor &rhs, const PermutationAlgebra<int,T> &pa, bool inverse=false, T alpha=1, T beta=0, bool conjugate_by_parity=false); //general (not optimally efficient) symmetrizers, antisymmetrizers etc. acting on the flattened index list:
void apply_permutation_algebra(const NRVec<Tensor> &rhsvec, const PermutationAlgebra<int,T> &pa, bool inverse=false, T alpha=1, T beta=0); //avoids explicit outer product but not vectorized, rather inefficient
// this *=beta; for I over this: this(I) += alpha * sum_P c_P rhs(P(I))
// PermutationAlgebra can represent e.g. general_antisymmetrizer in Kucharski-Bartlett notation, or Grassmann products building RDM from cumulants