Make mmf a parameter (proportion of mixing of two distributions)

epde/globals.py

@@ -98,7 +98,10 @@ def init_verbose(plot_DE_solutions : bool = False, show_iter_idx : bool = True,
                              show_iter_stats, show_warnings)
 
 
-def init_symnet_pool(max_factors_in_term, families, u, derivs, shape, names, grids, max_deriv_order):
+def init_symnet_pool(max_factors_in_term, families, u, derivs, shape, names, grids, max_deriv_order, mmf):
     global sympool
-    sympool = pool_terms.PoolTerms(max_factors_in_term, families)
+    sympool = pool_terms.PoolTerms(max_factors_in_term, families, mmf=mmf)
     sympool.set_initial_distr(u, derivs, shape, names, families=families, grids=grids, max_deriv_order=max_deriv_order)
+    # temporary_cf = [0.3, 1.5, 0.3, 0.3, 1.5, 0.3, 1.5, 0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,]
+    # for i, key in enumerate(sympool.pool_dict.keys()):
+    #     sympool.pool_dict[key] = temporary_cf[i]

epde/interface/interface.py

@@ -503,7 +503,7 @@ def set_preprocessor(self, preprocessor_pipeline: PreprocessingPipe = None,
 
     def create_pool(self, data: Union[np.ndarray, list, tuple], variable_names=['u',],
                     derivs=None, max_deriv_order=1, additional_tokens=[],
-                    data_fun_pow: int = 1, max_factors_in_term=1):
+                    data_fun_pow: int = 1, max_factors_in_term=1, mmf=2.4):
         self.pool_params = {'variable_names' : variable_names, 'max_deriv_order' : max_deriv_order,
                             'additional_tokens' : [family.token_family.ftype for family in additional_tokens]}
         assert (isinstance(derivs, list) and isinstance(derivs[0], np.ndarray)) or derivs is None
@@ -556,7 +556,7 @@ def create_pool(self, data: Union[np.ndarray, list, tuple], variable_names=['u',
         grids = [self.cache[0].memory_default.get('0'), self.cache[0].memory_default.get('1')]
         global_var.init_symnet_pool(max_factors_in_term, self.pool.families,
                                     entry.data_tensor, entry.derivatives, entry.data_tensor.shape, entry.names,
-                                    grids=grids, max_deriv_order=max_deriv_order)
+                                    grids=grids, max_deriv_order=max_deriv_order, mmf=mmf)
         print(f'The cardinality of defined token pool is {self.pool.families_cardinality()}')
         print(f'Among them, the pool contains {self.pool.families_cardinality(meaningful_only=True)}')
 
@@ -577,7 +577,7 @@ def saved_derivaties(self):
 
     def fit(self, data: Union[np.ndarray, list, tuple], equation_terms_max_number=6,
             equation_factors_max_number=1, variable_names=['u',], eq_sparsity_interval=(1e-4, 2.5),
-            derivs=None, max_deriv_order=1, additional_tokens=[], data_fun_pow: int = 1):
+            derivs=None, max_deriv_order=1, additional_tokens=[], data_fun_pow: int = 1, mmf: int = 2.4):
         """
         Fit epde search algorithm to obtain differential equations, describing passed data.
 
@@ -634,7 +634,8 @@ def fit(self, data: Union[np.ndarray, list, tuple], equation_terms_max_number=6,
                                  derivs=derivs, max_deriv_order=max_deriv_order,
                                  additional_tokens=additional_tokens,
                                  data_fun_pow=data_fun_pow,
-                                 max_factors_in_term=equation_factors_max_number)
+                                 max_factors_in_term=equation_factors_max_number,
+                                 mmf=mmf)
 
         self.optimizer_init_params['population_instruct'] = {"pool": self.pool, "terms_number": equation_terms_max_number,
                                                              "max_factors_in_term": equation_factors_max_number,

epde/operators/multiobjective/variation.py

@@ -202,11 +202,10 @@ def get_equation_cross_distr(self, equation, start_idx):
         for i in range(start_idx, len(equation.structure)):
             sym_term = to_symbolic(equation.structure[i])
             importance_coeffs[sym_term] = global_var.sympool.pool_dict.get(sym_term)
-        cross_distr = get_cross_distr(importance_coeffs, start_idx, len(equation.structure))
+        cross_distr = get_cross_distr(importance_coeffs, start_idx, len(equation.structure), global_var.sympool.mmf)
         return cross_distr
 
 
-
 class EquationExchangeCrossover(CompoundOperator):
     key = 'EquationExchangeCrossover'
 

epde/structure/main_structures.py

@@ -335,11 +335,11 @@ def set_mutation_probability(self):
         for tsym in allow_tsym:
             allow_csym.append(global_var.sympool.pool_dict.get(tsym))
 
-        eq_prob_ls = self._set_probabilities(allow_csym, draw_prob=False)
+        eq_prob_ls = self._set_probabilities(allow_csym, draw_prob=False, mmf=global_var.sympool.mmf)
         return allow_tsym, eq_prob_ls
 
     @staticmethod
-    def _set_probabilities(csym_pool_ls, mmf=2.4, draw_prob=False):
+    def _set_probabilities(csym_pool_ls, mmf=1.4, draw_prob=False):
         def draw_probabilities(final_probabilities, mmf, smoothing_factor, title):
             final_probabilities = final_probabilities.copy()
             final_probabilities.sort()
@@ -348,15 +348,13 @@ def draw_probabilities(final_probabilities, mmf, smoothing_factor, title):
             fig, ax = plt.subplots(figsize=(16, 8))
             ax.set_ylim(0, max(final_probabilities) + 0.01)
             sns.barplot(x=np.arange(len(final_probabilities)), y=final_probabilities, orient="v", ax=ax)
-            # plt.title(f"Smoothing factor: {smoothing_factor:.3f}, mmf: {mmf:.1f}")
             plt.grid()
             # plt.show()
             plt.yticks(fontsize=50)
             plt.savefig(f'individual_distr_{title}.png', transparent=True)
 
         csym_arr = np.fabs(np.array(csym_pool_ls))
-        # draw_probabilities(csym_arr, mmf, smoothing_factor, "before")
-        if np.max(csym_arr) / np.min(csym_arr) > 2.6:
+        if np.max(csym_arr) / np.min(csym_arr) > mmf:
             min_max_coeff = mmf * np.min(csym_arr) - np.max(csym_arr)
             smoothing_factor = min_max_coeff / (min_max_coeff - (mmf - 1) * np.average(csym_arr))
             uniform_csym = np.array([np.sum(csym_arr) / len(csym_arr)] * len(csym_arr))

symnet/initcoefficients.py

@@ -136,7 +136,7 @@ def get_csym_tsym(u, derivs, shape, input_names, pool_names, additional_tokens=N
     """
     # TODO: SymNet имеет 4 todo (+, pool_terms, preproc_input)
 
-    # TODO: если в левой части e.g. d^2u/dx2^2, то как получить в правой слагаемое d^2u/dx2^2 * u?
+    # TODO: если в левой части e.g. d^2u/dx1^2, то как получить в правой слагаемое d^2u/dx1^2 * u?
 
     left_pool = get_left_pool(max_deriv_order)
     model, left_side_name = select_model(input_names, left_pool, u, derivs, shape, additional_tokens)

symnet/pool_terms.py

@@ -6,11 +6,12 @@
 
 
 class PoolTerms:
-    def __init__(self, max_factors_in_term, families):
+    def __init__(self, max_factors_in_term, families, mmf):
 
         self.pool_sym_dict = None
         self.pool_sym_ls = None
         self.pool_dict = None
+        self.mmf = mmf
 
         token_ls = []
         for family in families:

symnet/preproc_output.py

@@ -57,8 +57,7 @@ def to_symbolic(term):
         return Symbol(str(term.cache_label[0]))
 
 
-def get_cross_distr(custom_cross_prob, start_idx, end_idx_exclude):
-    mmf = 2.4
+def get_cross_distr(custom_cross_prob, start_idx, end_idx_exclude, mmf):
     values = list(custom_cross_prob.values())
     csym_arr = np.fabs(np.array(values))