diff options
author | Michael Schneeberger <michael.schneeberger@fhnw.ch> | 2022-03-16 13:09:00 +0100 |
---|---|---|
committer | Michael Schneeberger <michael.schneeberger@fhnw.ch> | 2022-03-16 13:09:00 +0100 |
commit | f0981e0d4c85ddb8527501e77ee44dcbd86bdfea (patch) | |
tree | 79340a67199bc942004dc51a4d4ee2e8e2d8ab00 | |
parent | add inequality constraint and KKT conditions (diff) | |
download | polymatrix-f0981e0d4c85ddb8527501e77ee44dcbd86bdfea.tar.gz polymatrix-f0981e0d4c85ddb8527501e77ee44dcbd86bdfea.zip |
bugfixes in KKT conditions
-rw-r--r-- | polymatrix/init/initoptimization.py | 20 | ||||
-rw-r--r-- | polymatrix/init/initoptimizationstate.py | 10 | ||||
-rw-r--r-- | polymatrix/mixins/optimizationmixin.py | 524 | ||||
-rw-r--r-- | polymatrix/mixins/optimizationstatemixin.py | 6 | ||||
-rw-r--r-- | polymatrix/polysolver.py | 30 | ||||
-rw-r--r-- | polymatrix/sympyutils.py | 8 | ||||
-rw-r--r-- | test_polymatrix/test_polymatrix.py | 290 |
7 files changed, 588 insertions, 300 deletions
diff --git a/polymatrix/init/initoptimization.py b/polymatrix/init/initoptimization.py index c08c71f..653d248 100644 --- a/polymatrix/init/initoptimization.py +++ b/polymatrix/init/initoptimization.py @@ -1,13 +1,25 @@ from polymatrix.impl.optimizationimpl import OptimizationImpl +from polymatrix.init.initoptimizationstate import init_optimization_state from polymatrix.optimizationstate import OptimizationState def init_optimization( - state: OptimizationState, - equality_constraints: dict[int, dict[tuple[int, int], float]], - inequality_constraints: dict[int, dict[tuple[int, int], float]], - auxillary_equality: dict[int, dict[tuple[int, int], float]] + n_var: int, + equality_constraints: dict[int, dict[tuple[int, int], float]] = None, + inequality_constraints: dict[int, dict[tuple[int, int], float]] = None, + auxillary_equality: dict[int, dict[tuple[int, int], float]] = None, ): + state = init_optimization_state(n_var=n_var) + + if equality_constraints is None: + equality_constraints = {} + + if inequality_constraints is None: + inequality_constraints = {} + + if auxillary_equality is None: + auxillary_equality = {} + return OptimizationImpl( state=state, equality_constraints=equality_constraints, diff --git a/polymatrix/init/initoptimizationstate.py b/polymatrix/init/initoptimizationstate.py index e6e323d..372e225 100644 --- a/polymatrix/init/initoptimizationstate.py +++ b/polymatrix/init/initoptimizationstate.py @@ -4,9 +4,15 @@ from polymatrix.mixins.polymatrixmixin import PolyMatrixMixin def init_optimization_state( n_var: int, - n_param: int, - offset_dict: dict[tuple[PolyMatrixMixin, int], int], + n_param: int = None, + offset_dict: dict[tuple[PolyMatrixMixin, int], int] = None, ): + if n_param is None: + n_param = 0 + + if offset_dict is None: + offset_dict = {} + return OptimizationStateImpl( n_var=n_var, n_param=n_param, diff --git a/polymatrix/mixins/optimizationmixin.py b/polymatrix/mixins/optimizationmixin.py index d9ea263..e44a36f 100644 --- a/polymatrix/mixins/optimizationmixin.py +++ b/polymatrix/mixins/optimizationmixin.py @@ -5,6 +5,7 @@ import functools import itertools import more_itertools +from sympy import Equivalent from polymatrix.mixins.optimizationpipeopmixin import OptimizationPipeOpMixin from polymatrix.mixins.optimizationstatemixin import OptimizationStateMixin @@ -21,7 +22,7 @@ class OptimizationMixin(abc.ABC): @property @abc.abstractmethod - def equality_constraints(self) -> dict[int, dict[tuple[int, int], float]]: + def equality_constraints(self) -> dict[int, dict[tuple[int, tuple[int, ...], tuple[int, ...]], float]]: ... @property @@ -38,39 +39,43 @@ class OptimizationMixin(abc.ABC): def n_var(self) -> int: return self.state.n_var - @property - def n_equality_constraints(self) -> int: - if len(self.equality_constraints) == 0: - return 0 + # @property + # def n_equality_constraints(self) -> int: + # if len(self.equality_constraints) == 0: + # return 0 - return max(eq for eq, _ in self.equality_constraints.items()) + 1 + # return max(eq for eq, _ in self.equality_constraints.items()) + 1 - @property - def n_inequality_constraints(self) -> int: - if len(self.inequality_constraints) == 0: - return 0 + # @property + # def n_inequality_constraints(self) -> int: + # if len(self.inequality_constraints) == 0: + # return 0 - return max(eq for eq, _ in self.inequality_constraints.items()) + 1 - # return max(key[0] for _, d_data in self.inequality_constraints.items() for key, _ in d_data.items()) + 1 + # return max(eq for eq, _ in self.inequality_constraints.items()) + 1 + # # return max(key[0] for _, d_data in self.inequality_constraints.items() for key, _ in d_data.items()) + 1 - @property - def n_auxillary_equality(self) -> int: - if len(self.auxillary_equality) == 0: - return 0 + # @property + # def n_auxillary_equality(self) -> int: + # if len(self.auxillary_equality) == 0: + # return 0 - return max(eq for eq, _ in self.auxillary_equality.items()) + 1 - # return max(key[0] for _, d_data in self.auxillary_equality.items() for key, _ in d_data.items()) + 1 + # return max(eq for eq, _ in self.auxillary_equality.items()) + 1 + # # return max(key[0] for _, d_data in self.auxillary_equality.items() for key, _ in d_data.items()) + 1 # def pipe(self, *operators: OptimizationPipeOpMixin): # return functools.reduce(lambda obs, op: op(obs), operators, self) def add_equality_constraints(self, expr: tuple[tuple[PolyMatrixMixin, ...], ...]): + for term in expr: + for left, right in itertools.pairwise(term): + assert left.shape[1] == right.shape[0], f'{left} and {right} do not match' + all_polymats = tuple(polymat for term in expr for polymat in term) # update offsets with unseen polymats state = self.state.update_offsets(all_polymats) - eq_constr_buffer = collections.defaultdict(lambda: collections.defaultdict(float)) + equality_constraint = collections.defaultdict(float) for term in expr: @@ -85,36 +90,36 @@ class OptimizationMixin(abc.ABC): # n_var = 2, degree = 3 # cominations = [(0, 0, 0), (0, 0, 1), (0, 1, 1), (1, 1, 1)] for combination in itertools.combinations_with_replacement(range(self.n_var), total_degree): - for x_monom in more_itertools.distinct_permutations(combination): + for x_monomial in more_itertools.distinct_permutations(combination): def acc_func(acc, v): last, _ = acc new = last + v - return new, x_monom[last:new] + return new, x_monomial[last:new] # monom=(1, 0, 1) -> monom1=(x2, x1), monom2=(x2) # the monomial either refers to a free parameter, or to a value in the substitution table - p_monoms = list(monom for _, monom in itertools.accumulate(degrees, acc_func, initial=(0, None)))[1:] + rel_x_monomials = list(monom for _, monom in itertools.accumulate(degrees, acc_func, initial=(0, None)))[1:] def non_increasing(seq): return all(y <= x for x, y in zip(seq, seq[1:])) # (1,0) -> x2*x1 instead of (0,1)->x1*x2 - if all(non_increasing(monom) for monom in p_monoms): + if all(non_increasing(rel_x_monomial) for rel_x_monomial in rel_x_monomials): - col_defaults = [monomial_to_index(self.n_var, monom) for monom in p_monoms] + col_defaults = [monomial_to_index(self.n_var, monom) for monom in rel_x_monomials] n_rows_col = itertools.chain((m.shape[0] for m in term), (term[-1].shape[1],)) for curr_rows_col in itertools.product(*[range(e) for e in n_rows_col]): - def gen_re_indexing(): - for m, degree, (poly_row, poly_col), monom, col_default, subs, offset in zip( - term, degrees, itertools.pairwise(curr_rows_col), p_monoms, col_defaults, d_subs, d_offsets): + def gen_re_indexing(): #term=term, degrees=degrees, curr_rows_col=curr_rows_col, p_monoms=p_monoms, col_defaults=col_defaults, d_subs=d_subs, d_offsets=d_offsets): + for polymat, degree, (poly_row, poly_col), rel_x_monomial, col_default, subs, offset in zip( + term, degrees, itertools.pairwise(curr_rows_col), rel_x_monomials, col_defaults, d_subs, d_offsets): - if m.re_index is not None: - re_index = m.re_index(degree, poly_row, poly_col, monom) + if polymat.re_index is not None: + re_index = polymat.re_index(degree, poly_row, poly_col, rel_x_monomial) else: re_index = None @@ -138,11 +143,10 @@ class OptimizationMixin(abc.ABC): if subs_val is None: # the coefficient is selected after the reindexing - - row = new_poly_row + new_poly_col * m.shape[0] + row = new_poly_row + new_poly_col * polymat.shape[0] # linearize parameter matrix - p_index = int(offset[0] + row + col * m.shape[0] * m.shape[1]) + p_index = int(offset[0] + row + col * polymat.shape[0] * polymat.shape[1]) assert p_index < offset[1], f'{p_index} is bigger than {offset[1]}' @@ -164,25 +168,30 @@ class OptimizationMixin(abc.ABC): continue poly_row = data[0][0] - p_monom = tuple(d[2] for d in data if d[4] is None) - degree = len(p_monom) - - eq_constr_buffer[degree][poly_row, x_monom, p_monom] += value + p_monomial = tuple(sorted(d[2] for d in data if d[4] is None)) + x_monomial_sorted = tuple(sorted(x_monomial)) - # assign equations - rows_perm_set = set((eq_idx, perm) for eq_tuple_degree in eq_constr_buffer.values() for (eq_idx, perm, _) in eq_tuple_degree.keys()) - eq_to_rows = {eq: idx for idx, eq in enumerate(rows_perm_set)} + equality_constraint[poly_row, x_monomial_sorted, p_monomial] += value - eq_constr = collections.defaultdict(list) + # eq_constr = collections.defaultdict(list) - for degree, d_data in eq_constr_buffer.items(): - for (eq_idx, perm, p_monoms), val in d_data.items(): - row = eq_to_rows[(eq_idx, perm)] - eq_constr[row] += ((p_monoms, val),) + # for degree, d_data in eq_constr_buffer.items(): + # for (eq_idx, perm, p_monoms), val in d_data.items(): + # row = eq_to_rows[(eq_idx, perm)] + # eq_constr[row] += ((p_monoms, val),) # eq_data = dict(gen_eq_data()) - return dataclasses.replace(self, state=state, equality_constraints=eq_constr) + if len(self.equality_constraints) == 0: + constraint_index = 0 + else: + constraint_index = max(self.equality_constraints.keys()) + 1 + + return dataclasses.replace( + self, + state=state, + equality_constraints=self.equality_constraints | {constraint_index: equality_constraint}, + ) def add_inequality_constraints(self, expr): # all_polymats = tuple(polymat for term in expr for polymat in term) @@ -195,13 +204,11 @@ class OptimizationMixin(abc.ABC): aux_eq_buffer = collections.defaultdict(list) ineq_constr_buffer = collections.defaultdict(list) - ineq_idx = self.n_inequality_constraints - aux_eq_idx = self.n_auxillary_equality + ineq_idx = len(self.inequality_constraints) + aux_eq_idx = len(self.auxillary_equality) - for degree in polymat.degrees: - # print(f'{degree=}') - - assert degree != 0 + for degree in polymat.degrees: + # assert degree != 0 assert degree % 2 == 0 vec = tuple(itertools.combinations_with_replacement(range(state.n_var), int(degree/2))) @@ -209,20 +216,22 @@ class OptimizationMixin(abc.ABC): # introduce auxillary parameters offset_x = state.n_param - # print(f'{offset_x=}') n_param_added = int(n_square_mat*(n_square_mat-1)/2) state = state.update_param(n_param_added) - # print(f'{n_param_added=}') offset_a, end_idx = state.offset_dict[(polymat, degree)] def get_param_index_of_poly_mat(row, col): - x_monom = vec[row] + vec[col] - reindex = polymat.re_index(degree, 1, 1, x_monom) + x_monomial = tuple(sorted(vec[row] + vec[col], reverse=True)) + + if polymat.re_index is None: + reindex = None + else: + reindex = polymat.re_index(degree, 1, 1, x_monomial) if reindex is None: - n_monom, factor = x_monom, 1.0 + n_monom, factor = x_monomial, 1.0 else: _, _, n_monom, factor = polymat.re_index(degree, 1, 1, vec[row] + vec[col]) @@ -241,42 +250,46 @@ class OptimizationMixin(abc.ABC): for k in range(n_square_mat): # f in f-v^T@x-r^2 - monom, factor = get_param_index_of_poly_mat(k, k) - # ineq_constr_buffer[len(monom)][(ineq_idx, monom)] += factor - ineq_constr_buffer[ineq_idx] += ((monom, factor),) - - # print(ineq_constr_buffer) + p_monomial, factor = get_param_index_of_poly_mat(k, k) + ineq_constr_buffer[ineq_idx] += ((p_monomial, factor),) for row in range(k): # v^T@x in f-v^T@x-r^2 - monom, factor = get_param_index_of_poly_mat(k, row) - # ineq_constr_buffer[len(monom) + 1][(ineq_idx, monom + (offset_x + row,))] -= factor - ineq_constr_buffer[ineq_idx] += ((monom + (offset_x + row,), -factor),) - - # print(f'{k}: {offset_x + row=}') + p_monomial, factor = get_param_index_of_poly_mat(k, row) + ineq_constr_buffer[ineq_idx] += ((p_monomial + (offset_x + row,), -factor),) for col in range(k): # P@x in P@x-v - monom, factor = get_param_index_of_poly_mat(row, col) - # aux_eq_buffer[len(monom) + 1][(aux_eq_idx+row, monom + (offset_x + col,))] += factor - aux_eq_buffer[aux_eq_idx] += ((monom + (offset_x + col,), factor),) + p_monomial, factor = get_param_index_of_poly_mat(row, col) + aux_eq_buffer[aux_eq_idx] += ((p_monomial + (offset_x + col,), factor),) # -v in P@x-v - monom, factor = get_param_index_of_poly_mat(row, k) - # aux_eq_buffer[len(monom)][(aux_eq_idx+row, monom)] -= factor - aux_eq_buffer[aux_eq_idx] += ((monom, -factor),) + p_monomial, factor = get_param_index_of_poly_mat(row, k) + aux_eq_buffer[aux_eq_idx] += ((p_monomial, -factor),) aux_eq_idx += 1 ineq_idx += 1 offset_x += k - # print(f'{ineq_idx - self.n_inequality_constraints + aux_eq_idx - self.n_auxillary_equality=}') - # print(f'{aux_eq_idx - self.n_auxillary_equality=}') - return dataclasses.replace(self, inequality_constraints=ineq_constr_buffer, auxillary_equality=aux_eq_buffer, state=state) + def get_equality_constraints(self): + offset = 0 + equality_constraints = collections.defaultdict(list) + + for _, data in self.equality_constraints.items(): + equation_set = set((poly_row, x_monomial) for (poly_row, x_monomial, _) in data.keys()) + to_eq_index_dict = {eq: offset + idx for idx, eq in enumerate(equation_set)} + offset += len(equation_set) + + for (poly_row, x_monomial, p_monomial), val in data.items(): + eq_index = to_eq_index_dict[(poly_row, x_monomial)] + equality_constraints[eq_index] += ((p_monomial, val),) + + return equality_constraints + def minimize(self, cost_func=None): """ - assume sum of squares cost function on variables x @@ -293,17 +306,24 @@ class OptimizationMixin(abc.ABC): state = self.state - n_equality_constraints = self.n_equality_constraints - n_inequality_constraints = self.n_inequality_constraints + equality_constraints = self.get_equality_constraints() + + n_equality_constraints = len(equality_constraints) + + # if len(self.inequality_constraints) == 0: + # n_inequality_constraints = 0 + # else: + # n_inequality_constraints = max(eq for eq, _ in self.inequality_constraints.items()) + 1 + n_inequality_constraints = len(self.inequality_constraints) # remove unused parameters # ------------------------ - monom_update_reverse = set(m for monom_vals in itertools.chain(self.equality_constraints.values(), self.inequality_constraints.values(), self.auxillary_equality.values()) for monom, _ in monom_vals for m in monom) - # print(tuple(p for p in param_indices if p not in monom_update_reverse)) - monom_update = {monom: idx for idx, monom in enumerate(sorted(monom_update_reverse))} + # find used parameters + used_param_indices = set(param_idx for monomial_value_pairs in itertools.chain(equality_constraints.values(), self.inequality_constraints.values(), self.auxillary_equality.values()) for monomial, _ in monomial_value_pairs for param_idx in monomial) + param_update = {monom: idx for idx, monom in enumerate(sorted(used_param_indices))} - assert max(monom_update_reverse) == state.n_param - 1, f'{max(monom_update_reverse)=} is not {state.n_param - 1=}' + assert max(used_param_indices) == state.n_param - 1, f'{max(used_param_indices)=} is not {state.n_param - 1=}' # param_indices = tuple(start + idx for start, end in state.offset_dict.values() for idx in range(end - start)) # print(tuple(m for m in monom_update_reverse if m not in param_indices)) @@ -311,20 +331,18 @@ class OptimizationMixin(abc.ABC): # the variables x are assumed to be the parameters of all registered polynomial matrices # todo: this would later come from a specific polynomial matrices - param_indices = tuple(monom_update[start + idx] for start, end in state.offset_dict.values() for idx in range(end - start) if start + idx in monom_update_reverse) + param_indices = tuple(param_update[start + idx] for start, end in state.offset_dict.values() for idx in range(end - start) if start + idx in used_param_indices) - equality_constraints = tuple((key, tuple((tuple(monom_update[m] for m in monom), val) for monom, val in monom_val)) for key, monom_val in self.equality_constraints.items()) - inequality_constraints = tuple((key, tuple((tuple(monom_update[m] for m in monom), val) for monom, val in monom_val)) for key, monom_val in self.inequality_constraints.items()) - auxillary_equality = tuple((key, tuple((tuple(monom_update[m] for m in monom), val) for monom, val in monom_val)) for key, monom_val in self.auxillary_equality.items()) + equality_constraints = tuple((eq_index, tuple((tuple(param_update[m] for m in monom), val) for monom, val in monomial_value_pair)) for eq_index, monomial_value_pair in equality_constraints.items()) + inequality_constraints = tuple((key, tuple((tuple(param_update[m] for m in monom), val) for monom, val in monom_val)) for key, monom_val in self.inequality_constraints.items()) + auxillary_equality = tuple((key, tuple((tuple(param_update[m] for m in monom), val) for monom, val in monom_val)) for key, monom_val in self.auxillary_equality.items()) # introduce variable nu for each equality/inequality - idx_nu = len(monom_update_reverse) + idx_nu = len(used_param_indices) idx_lambda = idx_nu + n_equality_constraints idx_r1 = idx_lambda + n_inequality_constraints idx_r2 = idx_r1 + n_inequality_constraints - - # total_idx = idx_r2 + n_inequality_constraints - # print(f'{total_idx=}') + idx_dx = [idx_r2 + n_inequality_constraints] eq_buffer = collections.defaultdict(list) @@ -337,77 +355,315 @@ class OptimizationMixin(abc.ABC): current_eq_offset += n_equality_constraints assert max(eq_buffer.keys()) == current_eq_offset - 1 - # > inequality constraints - r1 + # > inequality constraints - r1^2 for eq, monom_val_list in inequality_constraints: - eq_buffer[current_eq_offset + eq] += monom_val_list + (((idx_r1 + eq,), -1),) + eq_buffer[current_eq_offset + eq] += monom_val_list + ((2 * (idx_r1 + eq,), -1),) current_eq_offset += n_inequality_constraints # assert max(eq_buffer.keys()) < current_eq_offset assert max(eq_buffer.keys()) == current_eq_offset - 1 - # > x + nu * equality constraint + lambda * inequality constraint - # --------------------------------------------------------------- + # assert max(eq_buffer.keys()) < current_eq_offset + n_inequality_constraints, f'{max(eq_buffer.keys())=} is bigger than {current_eq_offset + n_inequality_constraints=}' - for index, param in enumerate(param_indices): - eq_buffer[current_eq_offset + index] += (((param,), 1),) + dx_dict = collections.defaultdict(dict) - # assert max(eq_buffer.keys()) < current_eq_offset + n_inequality_constraints, f'{max(eq_buffer.keys())=} is bigger than {current_eq_offset + n_inequality_constraints=}' + for _, aux_eq_data in auxillary_equality: - # differentiate equality constraints for each variable x - # def gen_derivative(): - for ineq_constr_idx, eq_data in equality_constraints: - for p_monom, val in eq_data: + aux_variables = set(var for p_monomial, _ in aux_eq_data for var in p_monomial if var not in param_indices) + variables = set(var for p_monomial, _ in aux_eq_data for var in p_monomial if var in param_indices) + + # def gen_equation(): + # eq1 = (x, a d_a x) + # eq2 = (-1, a d_b x) + for var in variables: + + # print(f'{var=}') - p_monom_grp = collections.defaultdict(int) - for m in p_monom: - p_monom_grp[m] += 1 + # a x - b + # ((a, x), (-b,)) + for p_monomial, val in aux_eq_data: - for m, counter in p_monom_grp.items(): - if m in param_indices: + # print(f'{p_monomial=}') + + p_monomial_grp = dict(collections.Counter(p_monomial)) + + # d_a, d_b + # for var, counter in p_monomial_grp.items(): + # if var in param_indices: + + if var in p_monomial_grp: def generate_monom(): - for i_m, i_counter in p_monom_grp.items(): - if m is i_m: + # for each variable in the monomial + for i_var, i_counter in p_monomial_grp.items(): + if var is i_var: sel_counter = i_counter - 1 else: sel_counter = i_counter + # print(f'{(i_var, sel_counter)=}') + for _ in range(sel_counter): - yield i_m + yield i_var + + der_monomial = tuple(generate_monom()) + # if len(der_monomial): + # print(f'{der_monomial=}') + eq_buffer[current_eq_offset] += ((der_monomial, val * p_monomial_grp[var]),) + + for aux_var in aux_variables: + if aux_var in p_monomial_grp: + # create d_a x, d_b x, via the product rule + def generate_monom_2(): + # for each variable in the monomial + for i_var, i_counter in p_monomial_grp.items(): + if i_var is aux_var: + sel_counter = i_counter - 1 + else: + sel_counter = i_counter + + for _ in range(sel_counter): + yield i_var + + if var not in dx_dict: + dx_dict[var][aux_var] = idx_dx[0] + idx_dx[0] += 1 + + # product rule + yield dx_dict[var][aux_var] + + der_monomial_2 = tuple(generate_monom_2()) + # print(f'{der_monomial_2=}') + if len(der_monomial_2): + eq_buffer[current_eq_offset] += ((der_monomial_2, val * p_monomial_grp[aux_var]),) + + current_eq_offset += 1 + + # print(dx_dict) - eq_idx = param_indices.index(m) - der_monomial = tuple(generate_monom()) + (idx_nu + ineq_constr_idx,) - eq_buffer[current_eq_offset + eq_idx] += ((der_monomial, val * counter),) + # @dataclasses.dataclass + # class VariableIndexingState: + # index: int - for eq_constr_idx, eq_data in inequality_constraints: - for p_monom, val in eq_data: + # def increment(self): + # return dataclasses.replace(self, index=self.index + 1), self.index + + def differentiate(data, var, dx_dict): + eq_buffer = tuple() - p_monom_grp = collections.defaultdict(int) - for m in p_monom: - p_monom_grp[m] += 1 + # print(f'{var=}') + + for p_monomial, val in data: + + # count powers for each variable + p_monomial_grp = dict(collections.Counter(p_monomial)) + + # print(f'{p_monomial=}') + + # for var, counter in p_monomial_grp.items(): + if var in p_monomial_grp: + def generate_monom(): + for i_var, i_counter in p_monomial_grp.items(): + if var is i_var: + sel_counter = i_counter - 1 + else: + sel_counter = i_counter + + for _ in range(sel_counter): + yield i_var + + eq_buffer += (( + tuple(generate_monom()), + -val * p_monomial_grp[var], + ),) + + if var in dx_dict: + for aux_var in dx_dict[var]: + if aux_var in p_monomial_grp: + def generate_monom(): + for i_var, i_counter in p_monomial_grp.items(): + if i_var is aux_var: + sel_counter = i_counter - 1 + else: + sel_counter = i_counter + + for _ in range(sel_counter): + yield i_var + + yield dx_dict[var][aux_var] + + # der_monomial = tuple(generate_monom()) + + # if 1 < len(der_monomial): + # aux_eq_buffer += (( + # (der_monomial, 1), ((index_buffer[0],), -1) + # ),) + # variable_indices = (index_buffer[0],) + # else: + # variable_indices = der_monomial + + eq_buffer += (( + tuple(generate_monom()), + -val * p_monomial_grp[aux_var], + ),) + + # index_buffer[0] += 1 + + return eq_buffer + + def reduce_to_quadratic_equation(eq, index): + eq_buffer = tuple() + aux_eq_buffer = tuple() + index_buffer = [index] + + for monomial, val in eq: + n_aux = len(monomial) - 2 + + # print(f'{monomial=}') + # print(f'{n_aux=}') + + if 0 < n_aux: + eq_buffer += ((monomial[0:1] + (index_buffer[0],), val),) + + for i_aux in range(n_aux): + if i_aux == n_aux - 1: + aux_monom = monomial[i_aux+1:i_aux+3] + else: + aux_monom = monomial[i_aux+1:i_aux+2] + (index_buffer[0] + 1,) + + aux_eq_buffer += (( + (aux_monom, 1.0), ((index_buffer[0],), -1.0) + ),) + index_buffer[0] += 1 + else: + eq_buffer += ((monomial, val),) - for m, counter in p_monom_grp.items(): - if m in param_indices: - def generate_monom(): - for i_m, i_counter in p_monom_grp.items(): - if m is i_m: - sel_counter = i_counter - 1 - else: - sel_counter = i_counter - for _ in range(sel_counter): - yield i_m + return eq_buffer, aux_eq_buffer, index_buffer[0] + + + # > x + nu * equality constraint + lambda * inequality constraint + # --------------------------------------------------------------- + + for index, param in enumerate(param_indices): + eq_buffer[current_eq_offset + index] += (((param,), 1),) + + # differentiate equality constraints for each variable x + for eq_constr_idx, eq_data in equality_constraints: + for var in param_indices: + equation = differentiate(eq_data, var, dx_dict) + + equation = tuple((monomials + (idx_nu + eq_constr_idx,), val) for monomials, val in equation) + + equation, aux_equations, idx_dx[0] = reduce_to_quadratic_equation(equation, idx_dx[0]) + + if len(equation): + eq_idx = param_indices.index(var) + # print(f'{equation=}') + eq_buffer[current_eq_offset + eq_idx] += equation + + # # print(f'{aux_equations=}') + # for aux_eq in aux_equations: + # eq_buffer[next_current_eq_offset] = aux_eq + # next_current_eq_offset += 1 + + # for p_monomial, val in ineq_data: + + # p_monomial_grp = dict(collections.Counter(p_monomial)) + + # for var, counter in p_monomial_grp.items(): + # if var in param_indices: + # def generate_monom(): + # for i_m, i_counter in p_monomial_grp.items(): + # if var is i_m: + # sel_counter = i_counter - 1 + # else: + # sel_counter = i_counter + + # for _ in range(sel_counter): + # yield i_m + + # eq_idx = param_indices.index(var) + # der_monomial = tuple(generate_monom()) + (idx_nu + eq_constr_idx,) + # eq_buffer[current_eq_offset + eq_idx] += ((der_monomial, val * counter),) + + next_current_eq_offset = current_eq_offset + len(param_indices) + + # differentiate inequality constraints for each variable x + for ineq_constr_idx, ineq_data in inequality_constraints: + # aux_variables = set(var for p_monomial, _ in ineq_data for var in p_monomial if var not in param_indices) + # variables = set(var for p_monomial, _ in ineq_data for var in p_monomial if var in param_indices) + + for var in param_indices: + equation = differentiate(ineq_data, var, dx_dict) + + equation = tuple((monomials + (idx_lambda + ineq_constr_idx,), val) for monomials, val in equation) + + equation, aux_equations, idx_dx[0] = reduce_to_quadratic_equation(equation, idx_dx[0]) + # idx_dx[0] = index + + if len(equation): + eq_idx = param_indices.index(var) + # print(f'{equations=}') + eq_buffer[current_eq_offset + eq_idx] += equation + + # print(f'{aux_equations=}') + for aux_eq in aux_equations: + eq_buffer[next_current_eq_offset] = aux_eq + next_current_eq_offset += 1 + + # for p_monomial, val in ineq_data: + + # # count powers for each variable + # p_monomial_grp = dict(collections.Counter(p_monomial)) + + # # for var, counter in p_monomial_grp.items(): + # if var in p_monomial_grp: + # def generate_monom(): + # for i_var, i_counter in p_monomial_grp.items(): + # if var is i_var: + # sel_counter = i_counter - 1 + # else: + # sel_counter = i_counter + + # for _ in range(sel_counter): + # yield i_var + + # eq_idx = param_indices.index(var) + # der_monomial = tuple(generate_monom()) + (idx_lambda + ineq_constr_idx,) + # eq_buffer[current_eq_offset + eq_idx] += ((der_monomial, -val * counter),) + + # if var in dx_dict: + # for aux_var in dx_dict[var]: + # if aux_var in p_monomial_grp: + # def generate_monom(): + # for i_var, i_counter in p_monomial_grp.items(): + # if i_var is aux_var: + # sel_counter = i_counter - 1 + # else: + # sel_counter = i_counter + + # for _ in range(sel_counter): + # yield i_var + + # yield dx_dict[var][aux_var] + + # eq_idx = param_indices.index(var) + # der_monomial_2 = tuple(generate_monom()) + # eq_buffer[next_current_eq_offset] = ((der_monomial_2, 1), ((idx_dx[0],), -1)) + # next_current_eq_offset += 1 + + # eq_buffer[current_eq_offset + eq_idx] += (((idx_lambda + ineq_constr_idx, idx_dx[0]), -val * p_monomial_grp[aux_var]),) + # idx_dx[0] += 1 - eq_idx = param_indices.index(m) - der_monomial = tuple(generate_monom()) + (idx_lambda + eq_constr_idx,) - eq_buffer[current_eq_offset + eq_idx] += ((der_monomial, val * counter),) + # current_eq_offset += len(param_indices) + current_eq_offset = next_current_eq_offset - current_eq_offset += len(param_indices) # assert max(eq_buffer.keys()) < current_eq_offset, f'{max(eq_buffer.keys())=} is bigger than {current_eq_offset=}' assert max(eq_buffer.keys()) == current_eq_offset - 1 - # > lambda - r2 + # > lambda - r2^2 for idx in range(n_inequality_constraints): - eq_buffer[current_eq_offset + idx] += (((idx_lambda + idx,), 1), ((idx_r2 + idx,), -1)) + eq_buffer[current_eq_offset + idx] += (((idx_lambda + idx,), 1), (2 * (idx_r2 + idx,), -1)) current_eq_offset += n_inequality_constraints # assert max(eq_buffer.keys()) < current_eq_offset @@ -425,11 +681,11 @@ class OptimizationMixin(abc.ABC): for eq, data in auxillary_equality: eq_buffer[current_eq_offset + eq] += data - current_eq_offset += len(monom_update_reverse) - len(param_indices) + current_eq_offset += len(used_param_indices) - len(param_indices) # print(f'{current_eq_offset=}') # print(f'{len(monom_update_reverse) - len(param_indices)=}') assert max(eq_buffer.keys()) == current_eq_offset - 1, f'{max(eq_buffer.keys())} is not {current_eq_offset - 1}' # print(f'{current_eq_offset=}') - return eq_buffer, monom_update_reverse + return eq_buffer, param_update, current_eq_offset diff --git a/polymatrix/mixins/optimizationstatemixin.py b/polymatrix/mixins/optimizationstatemixin.py index ae688bb..731477c 100644 --- a/polymatrix/mixins/optimizationstatemixin.py +++ b/polymatrix/mixins/optimizationstatemixin.py @@ -48,9 +48,7 @@ class OptimizationStateMixin(abc.ABC): def update_offsets(self, polymats: tuple[PolyMatrixMixin]) -> 'OptimizationStateMixin': registered_polymats = set(polymat for polymat, _ in self.offset_dict.keys()) - # print(registered_polymats) parametric_polymats = set(p for p in polymats if not p.is_constant and p not in registered_polymats) - # print(parametric_polymats) if len(parametric_polymats) == 0: return self @@ -64,13 +62,15 @@ class OptimizationStateMixin(abc.ABC): # number of terms is given by the maximum index + 1 number_of_terms = int(polymat.shape[0] * polymat.shape[1] * (monomial_to_index(self.n_var, degree*(self.n_var-1,)) + 1)) + # print(f'{polymat=}, {number_of_terms=}, {polymat.shape[0] * polymat.shape[1]=}, {monomial_to_index(self.n_var, degree*(self.n_var-1,)) + 1=}') + yield (polymat, degree), number_of_terms param_key, num_of_terms = tuple(zip(*gen_n_param_per_struct())) cum_sum = tuple(itertools.accumulate((self.n_param,) + num_of_terms)) offset_dict = dict(zip(param_key, itertools.pairwise(cum_sum))) - return dataclasses.replace(self, offset_dict=offset_dict, n_param=cum_sum[-1]) + return dataclasses.replace(self, offset_dict=self.offset_dict | offset_dict, n_param=cum_sum[-1]) def update_param(self, n): return dataclasses.replace(self, n_param=self.n_param+n) diff --git a/polymatrix/polysolver.py b/polymatrix/polysolver.py index 24c7095..1f7bc8e 100644 --- a/polymatrix/polysolver.py +++ b/polymatrix/polysolver.py @@ -12,12 +12,11 @@ import time def substitude_x_add_a(data, a): def kron_eye_and_a(n, data, a, indices): - permutations = more_itertools.distinct_permutations(indices) - # permutations = set(itertools.permutations(indices)) + # permutations = more_itertools.distinct_permutations(indices) + permutations = set(itertools.permutations(indices)) if not scipy.sparse.issparse(data): - a_mat = np.reshape(a, (-1, 1)) - ops = (np.eye(n), a_mat) + ops = (np.eye(n), np.reshape(a, (-1, 1))) def acc_kron(perm): *_, last = itertools.accumulate((ops[idx] for idx in perm), lambda acc, v: np.kron(acc, v)) @@ -103,6 +102,8 @@ def substitude_x_add_a(data, a): yield key, d_data return dict(gen_subs_data()) + # return subs_data + def solve_poly_system(data, m): if isinstance(data[1], np.ndarray): @@ -267,17 +268,24 @@ def outer_smart_solve(data, n_iter=10, a_max=1.0, irange=None, idegree=None): def gen_a_err(): n = data[1].shape[0] - for idx_iter in range(n_iter): + for _ in range(n_iter): a_init = a_max * (2 * np.random.rand(n) - 1) - a, err = inner_smart_solve(data, irange=irange, idegree=idegree, a_init=a_init) + try: + a, err = inner_smart_solve(data, irange=irange, idegree=idegree, a_init=a_init) - subs_data = substitude_x_add_a(data, a[-1]) - sol = solve_poly_system(subs_data, 6) - error_index = np.max(np.abs(eval_solution(subs_data, sol))) + max(a[-1]) - # error_index = np.max(np.abs(eval_solution(subs_data, sol))) + subs_data = substitude_x_add_a(data, a[-1]) + sol = solve_poly_system(subs_data, 6) + error_index = np.max(np.abs(eval_solution(subs_data, sol))) + max(a[-1]) + # error_index = np.max(np.abs(eval_solution(subs_data, sol))) + + except: + print(f'nan error, continue') + # print(f'nan error for {a_init=}, continue') + # yield np.nan, a_init, np.nan + continue - # print(f'{error_index=}') + print(f'{error_index=}') yield error_index, a, err diff --git a/polymatrix/sympyutils.py b/polymatrix/sympyutils.py index 7413ba3..ac91b8f 100644 --- a/polymatrix/sympyutils.py +++ b/polymatrix/sympyutils.py @@ -55,23 +55,23 @@ def poly_to_data_coord(poly_list, x, degree = None): def poly_to_matrix(poly_list, x, power = None): data_coord_dict = poly_to_data_coord(poly_list, x, power) - n_free_symbols = len(x) + n_var = len(x) n_poly = len(poly_list) n_eq = sum((1 for row_eq in poly_list for e in row_eq)) def gen_power_mat(): if 0 not in data_coord_dict: - yield 0, np.zeros((n_free_symbols, 1)) + yield 0, np.zeros((n_var, 1)) # for all powers generate a matrix for degree, data_coord in data_coord_dict.items(): # empty matrix - shape = (n_eq, n_free_symbols**degree) + shape = (n_eq, n_var**degree) # fill matrix if len(data_coord) == 0: - yield np.zeros((len(poly_list), n_free_symbols**degree)) + yield np.zeros((len(poly_list), n_var**degree)) else: def gen_row_col_coord(): diff --git a/test_polymatrix/test_polymatrix.py b/test_polymatrix/test_polymatrix.py index f693df0..d8abeef 100644 --- a/test_polymatrix/test_polymatrix.py +++ b/test_polymatrix/test_polymatrix.py @@ -1,22 +1,55 @@ import unittest - -from polymatrix.polystruct import init_equation, init_poly_matrix -from polymatrix.utils import variable_to_index +from polymatrix.init.initoptimization import init_optimization +from polymatrix.init.initpolymatrix import init_poly_matrix +from polymatrix.optimization import Optimization +from polymatrix.polymatrix import PolyMatrix class TestPolyMatrix(unittest.TestCase): + # @staticmethod + # def assert_term_in_eq(result, degree, eq_idx, row_idx, value, monoms=None): + # if monoms is None: + # monoms = tuple() + + # assert degree in result.data, f'could not find {degree} in {result.data}' + # degree_terms = result.data[degree] + + # key = eq_idx, monoms, row_idx + # assert key in degree_terms, f'could not find {key} in {degree_terms}' + # # eq_terms = degree_terms[key] + # # assert row_idx in eq_terms, f'could not find {row_idx} in {eq_terms}' + # assert degree_terms[key] == value, f'value {degree_terms[key]} and {value} do not match' + + # @staticmethod + # def assert_term_in_eq(result, degree, row_idx, monoms, value): + + # assert degree in result.data, f'could not find {degree} in {result.data}' + # degree_terms = result.data[degree] + + # key = row_idx, variable_to_index(result.n_param, monoms) + # assert key in degree_terms, f'could not find {key} in {degree_terms}' + + # assert degree_terms[key] == value, f'value {degree_terms[key]} and {value} do not match' + @staticmethod - def assert_term_in_eq(result, degree, eq_idx, row_idx, value, monoms=None): - if monoms is None: - monoms = tuple() + def assert_term_in_eq( + problem: Optimization, + poly_row: int, + p_monomial: tuple[tuple[PolyMatrix, int, int], ...], + value: float, + x_monomial: tuple[int, ...] = None, + ): + if x_monomial is None: + x_monomial = tuple() + + offset_dict = problem.state.offset_dict + p_monomial = tuple(offset_dict[(polymat, degree)][0] + offset for polymat, degree, offset in p_monomial) - assert degree in result.terms, f'could not find {degree} in {result.terms}' - degree_terms = result.terms[degree] + equality_constraint = problem.equality_constraints[0] - key = eq_idx, monoms, row_idx - assert key in degree_terms, f'could not find {key} in {degree_terms}' - # eq_terms = degree_terms[key] - # assert row_idx in eq_terms, f'could not find {row_idx} in {eq_terms}' - assert degree_terms[key] == value, f'value {degree_terms[key]} and {value} do not match' + key = (poly_row, x_monomial, p_monomial) + assert key in equality_constraint, f'could not find {key} in {equality_constraint}' + + assert equality_constraint[key] == value, f'value {equality_constraint[key]} and {value} do not match' def test_param_matrix_param_d0_vector_degree_d0(self): """ @@ -25,50 +58,44 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 - mat = init_poly_matrix(degrees=(0,), shape=(n_var, n_var)) - vec = init_poly_matrix(degrees=(0,), shape=(n_var, 1)) + mat = init_poly_matrix(name='mat', degrees=(0,), shape=(n_var, n_var)) + vec = init_poly_matrix(name='vec', degrees=(0,), shape=(n_var, 1)) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - # a11 v11 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 0, - row_idx = (offset_dict[(mat, 0)], offset_dict[(vec, 0)]), + problem = problem, + poly_row = 0, + p_monomial = ((mat, 0, 0), (vec, 0, 0)), value = 1, ) # a12 v21 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 0, - row_idx = (offset_dict[(mat, 0)]+2, offset_dict[(vec, 0)]+1), + problem = problem, + poly_row = 0, + p_monomial = ((mat, 0, 2), (vec, 0, 1)), value = 1, ) # a21 v11 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 1, - row_idx = (offset_dict[(mat, 0)]+1, offset_dict[(vec, 0)]), + problem = problem, + poly_row = 1, + p_monomial = ((mat, 0, 1), (vec, 0, 0)), value = 1, ) # a22 v21 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 1, - row_idx = (offset_dict[(mat, 0)]+3, offset_dict[(vec, 0)]+1), + problem = problem, + poly_row = 1, + p_monomial = ((mat, 0, 3), (vec, 0, 1)), value = 1, ) @@ -79,33 +106,29 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 - mat = init_poly_matrix(degrees=(0,), shape=(n_var, n_var)) - vec = init_poly_matrix(degrees=(0,1), shape=(n_var, 1)) + mat = init_poly_matrix(name='mat', degrees=(0,), shape=(n_var, n_var)) + vec = init_poly_matrix(name='vec', degrees=(0, 1), shape=(n_var, 1)) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - # a11 v011 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 0, - row_idx = (offset_dict[(mat, 0)], offset_dict[(vec, 0)]), + problem = problem, + poly_row = 0, + p_monomial = ((mat, 0, 0), (vec, 0, 0)), value = 1, ) # a11 v011 self.assert_term_in_eq( - result = result, - degree = 2, - eq_idx = 0, - monoms=(0,), - row_idx = (offset_dict[(mat, 0)], offset_dict[(vec, 1)]), + problem = problem, + poly_row = 0, + x_monomial=(0,), + p_monomial = ((mat, 0, 0), (vec, 1, 0)), value = 1, ) @@ -116,32 +139,29 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 - mat = init_poly_matrix(degrees=(0,), shape=(n_var, n_var)) - vec = init_poly_matrix(subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) + mat = init_poly_matrix(name='mat', degrees=(0,), shape=(n_var, n_var)) + vec = init_poly_matrix(name='vec', subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict # a11 self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 0, - row_idx = (offset_dict[(mat, 0)],), + problem = problem, + poly_row = 0, + p_monomial = ((mat, 0, 0),), value = 1, ) # a21 self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 1, - row_idx = (offset_dict[(mat, 0)]+1,), + problem = problem, + poly_row = 1, + p_monomial = ((mat, 0, 1),), value = 1, ) @@ -151,45 +171,40 @@ class TestPolyMatrix(unittest.TestCase): """ n_var = 2 - - mat = init_poly_matrix(degrees=(0,), shape=(n_var, n_var)) - vec = init_poly_matrix(subs={1: {(0, 0, 0): 1, (0, 0, 1): 0, (1, 0, 0): 0, (1, 0, 1): 1}}, shape=(n_var, 1)) - - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + + mat = init_poly_matrix(name='mat', degrees=(0,), shape=(n_var, n_var)) + vec = init_poly_matrix(name='vec', subs={1: {(0, 0, 0): 1, (0, 0, 1): 0, (1, 0, 0): 0, (1, 0, 1): 1}}, shape=(n_var, 1)) + + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - # a11 self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 0, - monoms=(0,), - row_idx = (offset_dict[(mat, 0)],), + problem = problem, + poly_row = 0, + x_monomial=(0,), + p_monomial = ((mat, 0, 0),), value = 1, ) # a12 self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 0, - monoms=(1,), - row_idx = (offset_dict[(mat, 0)]+2,), + problem = problem, + poly_row = 0, + x_monomial=(1,), + p_monomial = ((mat, 0, 2),), value = 1, ) # a21 self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 1, - monoms=(0,), - row_idx = (offset_dict[(mat, 0)]+1,), + problem = problem, + poly_row = 1, + x_monomial=(0,), + p_monomial = ((mat, 0, 1),), value = 1, ) @@ -200,22 +215,19 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 - mat = init_poly_matrix(subs={0: {(0, 0, 0): 1, (0, 1, 0): 1, (1, 0, 0): 1, (1, 1, 0): 1}}, shape=(n_var, n_var)) - vec = init_poly_matrix(subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) + mat = init_poly_matrix(name='mat', subs={0: {(0, 0, 0): 1, (0, 1, 0): 1, (1, 0, 0): 1, (1, 1, 0): 1}}, shape=(n_var, n_var)) + vec = init_poly_matrix(name='vec', subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - self.assert_term_in_eq( - result = result, - degree = 0, - eq_idx = 0, - row_idx = tuple(), + problem = problem, + poly_row = 0, + p_monomial = tuple(), value = 2, ) @@ -233,33 +245,30 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 mat = init_poly_matrix( + name='mat', degrees=(0,), re_index=skew_symmetric, shape=(n_var, n_var), ) - vec = init_poly_matrix(subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) + vec = init_poly_matrix(name='vec', subs={0: {(0, 0, 0): 1, (1, 0, 0): 1}}, shape=(n_var, 1)) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 0, - row_idx = (offset_dict[(mat, 0)] + 2,), + problem = problem, + poly_row = 0, + p_monomial = ((mat, 0, 2),), value = 1, ) self.assert_term_in_eq( - result = result, - degree = 1, - eq_idx = 1, - row_idx = (offset_dict[(mat, 0)] + 2,), + problem = problem, + poly_row = 1, + p_monomial = ((mat, 0, 2),), value = -1, ) @@ -285,46 +294,43 @@ class TestPolyMatrix(unittest.TestCase): n_var = 2 mat = init_poly_matrix( + name='mat', subs={0: {(0, 0, 0): 1, (0, 1, 0): 1, (1, 0, 0): 1, (1, 1, 0): 1}}, shape=(n_var, n_var), ) vec = init_poly_matrix( + name='vec', degrees=(1,), re_index=gradient, shape=(n_var, 1), ) - eq = init_equation( - terms = [(mat, vec)], - n_var = n_var, + problem = init_optimization( + n_var=n_var, + ).add_equality_constraints( + expr=[(mat, vec)], ) - result = eq.create() - offset_dict = eq.offset_dict - self.assert_term_in_eq( - result = result, - degree = 1, - monoms=(0,), - eq_idx = 0, - row_idx = (offset_dict[(vec, 1)],), + problem = problem, + x_monomial=(0,), + poly_row = 0, + p_monomial = ((vec, 1, 0),), value = 2, ) self.assert_term_in_eq( - result = result, - degree = 1, - monoms=(1,), - eq_idx = 0, - row_idx = (offset_dict[(vec, 1)]+2,), + problem = problem, + x_monomial=(1,), + poly_row = 0, + p_monomial = ((vec, 1, 2),), value = 1, ) self.assert_term_in_eq( - result = result, - degree = 1, - monoms=(0,), - eq_idx = 1, - row_idx = (offset_dict[(vec, 1)]+2,), + problem = problem, + x_monomial=(0,), + poly_row = 1, + p_monomial = ((vec, 1, 2),), value = 1, ) |