Skip to content

GitLab

Commit 56bb43ce authored by ferhat's avatar ferhat
Browse files

- sparse_dll option works fine with feedback variables 

git-svn-id: https://www.dynare.org/svn/dynare/trunk@2851 ac1d8469-bf42-47a9-8791-bf33cf982152
parent 4b48b5ef
Expand all Hide whitespace changes
Inline Side-by-side
BlockTriangular.cc
View file @ 56bb43ce
......@@ -328,7 +328,7 @@ BlockTriangular::Allocate_Block(int size, int *count_Equ, int count_Block, Block
ModelBlock->Block_List[count_Block].Type = type;
ModelBlock->Block_List[count_Block].Nb_Recursives = recurs_Size;
ModelBlock->Block_List[count_Block].Temporary_InUse = new temporary_terms_inuse_type();
ModelBlock->Block_List[count_Block].Chaine_Rule_Derivatives = new chaine_rule_derivatives_type();
ModelBlock->Block_List[count_Block].Chain_Rule_Derivatives = new chain_rule_derivatives_type();
ModelBlock->Block_List[count_Block].Temporary_InUse->clear();
ModelBlock->Block_List[count_Block].Simulation_Type = SimType;
ModelBlock->Block_List[count_Block].Equation = (int *) malloc(ModelBlock->Block_List[count_Block].Size * sizeof(int));
......@@ -341,7 +341,7 @@ BlockTriangular::Allocate_Block(int size, int *count_Equ, int count_Block, Block
first_count_equ = *count_Equ;
tmp_var = (int *) malloc(size * sizeof(int));
tmp_endo = (int *) malloc((incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1) * sizeof(int));
tmp_other_endo = (int *) malloc(symbol_table.endo_nbr() * sizeof(int));
tmp_other_endo = (int *) malloc((incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1) * sizeof(int));
tmp_size = (int *) malloc((incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1) * sizeof(int));
tmp_size_other_endo = (int *) malloc((incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1) * sizeof(int));
tmp_size_exo = (int *) malloc((incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1) * sizeof(int));
......@@ -349,7 +349,7 @@ BlockTriangular::Allocate_Block(int size, int *count_Equ, int count_Block, Block
memset(tmp_size_other_endo, 0, (incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1)*sizeof(int));
memset(tmp_size, 0, (incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1)*sizeof(int));
memset(tmp_endo, 0, (incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1)*sizeof(int));
memset(tmp_other_endo, 0, symbol_table.endo_nbr()*sizeof(int));
memset(tmp_other_endo, 0, (incidencematrix.Model_Max_Lead + incidencematrix.Model_Max_Lag + 1)*sizeof(int));
nb_lead_lag_endo = 0;
Lag_Endo = Lead_Endo = Lag_Other_Endo = Lead_Other_Endo = Lag_Exo = Lead_Exo = 0;
......@@ -438,7 +438,7 @@ BlockTriangular::Allocate_Block(int size, int *count_Equ, int count_Block, Block
{
if (!tmp_variable_evaluated[j])
{
tmp_other_endo[j] = 1;
tmp_other_endo[incidencematrix.Model_Max_Lag + k]++;
tmp_nb_other_endo++;
}
if (k > 0 && k > Lead_Other_Endo)
......@@ -679,7 +679,7 @@ BlockTriangular::Free_Block(Model_Block *ModelBlock) const
delete ModelBlock->Block_List[blk].Temporary_Terms_in_Equation[i];
free(ModelBlock->Block_List[blk].Temporary_Terms_in_Equation);
delete (ModelBlock->Block_List[blk].Temporary_InUse);
delete ModelBlock->Block_List[blk].Chaine_Rule_Derivatives;
delete ModelBlock->Block_List[blk].Chain_Rule_Derivatives;
}
free(ModelBlock->Block_List);
free(ModelBlock);
......@@ -728,12 +728,13 @@ BlockTriangular::Equation_Type_determination(vector<BinaryOpNode *> &equations,
}
else
{
//vector<pair<int, NodeID> > List_of_Op_RHS;
//the equation could be normalized by a permutation of the rhs and the lhs
if (d_endo_variable == result.end()) //the equation is linear in the endogenous and could be normalized using the derivative
{
Equation_Simulation_Type = E_EVALUATE_S;
//cout << " gone normalized : ";
res = equations[eq]->normalizeLinearInEndoEquation(var, derivative->second);
res = equations[eq]->normalizeLinearInEndoEquation(var, derivative->second/*, List_of_Op_RHS*/);
/*res.second->writeOutput(cout, oMatlabDynamicModelSparse, temporary_terms);
cout << " done\n";*/
}
......@@ -854,10 +855,11 @@ BlockTriangular::Reduce_Blocks_and_type_determination(int prologue, int epilogue
return (Type);
}
map<pair<pair<int, int>, pair<pair<int, int>, int> >, int>
map<pair<pair<int, pair<int, int> >, pair<int, int> >, int>
BlockTriangular::get_Derivatives(Model_Block *ModelBlock, int blck)
{
map<pair<pair<int, int>, pair<pair<int, int>, int> >, int> Derivatives;
map<pair<pair<int, pair<int, int> >, pair<int, int> >, int> Derivatives;
Derivatives.clear();
int nb_endo = symbol_table.endo_nbr();
/*ModelBlock.Block_List[Blck].first_order_determinstic_simulation_derivatives = new*/
......@@ -876,11 +878,8 @@ BlockTriangular::get_Derivatives(Model_Block *ModelBlock, int blck)
cout << "varr=" << varr << " eqr=" << eqr << " lag=" << lag << "\n";*/
if(IM[varr+eqr*nb_endo])
{
/*if(eq<ModelBlock->Block_List[blck].Nb_Recursives and var<ModelBlock->Block_List[blck].Nb_Recursives)
{*/
bool OK = true;
map<pair<pair<int, int>, pair<pair<int, int>, int> >, int>::const_iterator its = Derivatives.find(make_pair(make_pair(eqr, eq), make_pair(make_pair(varr, var), lag)));
map<pair<pair<int, pair<int, int> >, pair<int, int> >, int>::const_iterator its = Derivatives.find(make_pair(make_pair(lag, make_pair(eq, var)), make_pair(eqr, varr)));
if(its!=Derivatives.end())
{
if(its->second == 2)
......@@ -890,20 +889,21 @@ BlockTriangular::get_Derivatives(Model_Block *ModelBlock, int blck)
if(OK)
{
if (ModelBlock->Block_List[blck].Equation_Type[eq] == E_EVALUATE_S and eq<ModelBlock->Block_List[blck].Nb_Recursives)
Derivatives[make_pair(make_pair(eqr, eq), make_pair(make_pair(varr, var), lag))] = 1;
//It's a normalized equation, we have to recompute the derivative using chain rule derivative function*/
Derivatives[make_pair(make_pair(lag, make_pair(eq, var)), make_pair(eqr, varr))] = 1;
else
Derivatives[make_pair(make_pair(eqr, eq),make_pair(make_pair(varr, var), lag))] = 0;
//It's a feedback equation we can use the derivatives
Derivatives[make_pair(make_pair(lag, make_pair(eq, var)), make_pair(eqr, varr))] = 0;
}
/*}
else if(eq<ModelBlock->Block_List[blck].Nb_Recursives and var<ModelBlock->Block_List[blck].Nb_Recursives)*/
if(var<ModelBlock->Block_List[blck].Nb_Recursives)
{
int eqs = ModelBlock->Block_List[blck].Equation[var];
for(int vars=ModelBlock->Block_List[blck].Nb_Recursives; vars<ModelBlock->Block_List[blck].Size; vars++)
{
int varrs = ModelBlock->Block_List[blck].Variable[vars];
if(Derivatives.find(make_pair(make_pair(eqs, var), make_pair(make_pair(varrs, vars), lag)))!=Derivatives.end())
Derivatives[make_pair(make_pair(eqr, eq),make_pair(make_pair(varrs, vars), lag))] = 2;
//A new derivative need to be computed using the chain rule derivative function (a feedback variable appear in a recursive equation)
if(Derivatives.find(make_pair(make_pair(lag, make_pair(var, vars)), make_pair(eqs, varrs)))!=Derivatives.end())
Derivatives[make_pair(make_pair(lag, make_pair(eq, vars)), make_pair(eqr, varrs))] = 2;
}
}
}
......
BlockTriangular.hh
View file @ 56bb43ce
......@@ -44,7 +44,7 @@ typedef vector<pair< int, int> > t_vtype;
typedef set<int> temporary_terms_inuse_type;
typedef vector<pair< pair<int, int>, pair<int, pair<int, int> > > > chaine_rule_derivatives_type;
typedef vector<pair< pair<int, pair<int, int> >, pair<int, int> > > chain_rule_derivatives_type;
//! For one lead/lag of one block, stores mapping of information between original model and block-decomposed model
struct IM_compact
......@@ -73,7 +73,7 @@ struct Block
//temporary_terms_type *Temporary_terms;
temporary_terms_inuse_type *Temporary_InUse;
IM_compact *IM_lead_lag;
chaine_rule_derivatives_type *Chaine_Rule_Derivatives;
chain_rule_derivatives_type *Chain_Rule_Derivatives;
int Code_Start, Code_Length;
};
......@@ -118,7 +118,7 @@ public:
//! Frees the Model structure describing the content of each block
void Free_Block(Model_Block* ModelBlock) const;
map<pair<pair<int, int>, pair<pair<int, int>, int> >, int> get_Derivatives(Model_Block *ModelBlock, int Blck);
map<pair<pair<int, pair<int, int> >, pair<int, int> >, int> get_Derivatives(Model_Block *ModelBlock, int Blck);
void Normalize_and_BlockDecompose_Static_0_Model(jacob_map &j_m, vector<BinaryOpNode *> &equations, t_etype &V_Equation_Type, map<pair<int, pair<int, int> >, NodeID> &first_order_endo_derivatives);
......
DynamicModel.cc
View file @ 56bb43ce
This diff is collapsed.
DynamicModel.hh
View file @ 56bb43ce
......@@ -77,8 +77,8 @@ private:
//! Temporary terms for the file containing parameters dervicatives
temporary_terms_type params_derivs_temporary_terms;
typedef map< pair< int, pair< int, int> >, NodeID> first_chaine_rule_derivatives_type;
first_chaine_rule_derivatives_type first_chaine_rule_derivatives;
typedef map< pair< int, pair< int, int> >, NodeID> first_chain_rule_derivatives_type;
first_chain_rule_derivatives_type first_chain_rule_derivatives;
//! Writes dynamic model file (Matlab version)
......@@ -110,6 +110,8 @@ private:
void computeTemporaryTermsOrdered(Model_Block *ModelBlock);
//! Write derivative code of an equation w.r. to a variable
void compileDerivative(ofstream &code_file, int eq, int symb_id, int lag, map_idx_type &map_idx) const;
//! Write chain rule derivative code of an equation w.r. to a variable
void compileChainRuleDerivative(ofstream &code_file, int eq, int var, int lag, map_idx_type &map_idx) const;
virtual int computeDerivID(int symb_id, int lag);
//! Get the type corresponding to a derivation ID
......@@ -120,8 +122,8 @@ private:
int getSymbIDByDerivID(int deriv_id) const throw (UnknownDerivIDException);
//! Compute the column indices of the dynamic Jacobian
void computeDynJacobianCols(bool jacobianExo);
//! Computes chaine rule derivatives of the Jacobian w.r. to endogenous variables
void computeChaineRuleJacobian(Model_Block *ModelBlock);
//! Computes chain rule derivatives of the Jacobian w.r. to endogenous variables
void computeChainRuleJacobian(Model_Block *ModelBlock);
//! Computes derivatives of the Jacobian w.r. to parameters
void computeParamsDerivatives();
//! Computes temporary terms for the file containing parameters derivatives
......@@ -137,8 +139,8 @@ private:
/*! Writes either (i+1,j+1) or [i+j*NNZDerivatives[1]] */
void hessianHelper(ostream &output, int row_nb, int col_nb, ExprNodeOutputType output_type) const;
//! Write chaine rule derivative of a recursive equation w.r. to a variable
void writeChaineRuleDerivative(ostream &output, int eq, int var, int lag, ExprNodeOutputType output_type, const temporary_terms_type &temporary_terms) const;
//! Write chain rule derivative of a recursive equation w.r. to a variable
void writeChainRuleDerivative(ostream &output, int eq, int var, int lag, ExprNodeOutputType output_type, const temporary_terms_type &temporary_terms) const;
public:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment