metacountregressor 0.1.315__tar.gz → 0.1.317__tar.gz

{metacountregressor-0.1.315/metacountregressor.egg-info → metacountregressor-0.1.317}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: metacountregressor
-Version: 0.1.315
+Version: 0.1.317
 Summary: Extensive Testing for Estimation of Data Count Models
 Home-page: https://github.com/zahern/CountDataEstimation
 Author: Zeke Ahern

{metacountregressor-0.1.315 → metacountregressor-0.1.317}/metacountregressor/metaheuristics.py

@@ -65,7 +65,7 @@ def _plot(x, y, z, xlabel=None, ylabel=None, zlabel=None, filename=None):
 def dict_mean(dict_list,
               ignore=None):
     if ignore is None:
-        ignore = ['AIC', 'layout', 'fixed_fit', 'rdm_fit', 'rdm_cor_fit', 'zi_fit', 'simple', 'pvalues']
+        ignore = ['AIC', 'layout', 'fixed_fit', 'rdm_fit', 'rdm_cor_fit', 'zi_fit', 'simple', 'pvalues', "CAIC"]
     mean_dict = {}
     if ignore is None:
         for key in dict_list[0].keys():
@@ -76,7 +76,13 @@ def dict_mean(dict_list,
         for key in dict_list[0].keys():
             if key in ignore:
                 continue
-            mean_dict[key] = sum(d[key] for d in dict_list) / len(dict_list)
+            # Filter out dictionaries that don't have the key
+            filtered_values = [d[key] for d in dict_list if key in d]
+            
+            if filtered_values:  # Ensure there are values to compute the mean
+                mean_dict[key] = sum(filtered_values) / len(filtered_values)
+            else:
+                mean_dict[key] = None  # Or handle missing data differently if needed
         return mean_dict
 
 
@@ -1198,7 +1204,7 @@ class SimulatedAnnealing(object):
             Temp3 = (Temp1 + Temp2) / (w_1 * Temp2 + w_2 * Temp1)
 
             self.best_energy = slns[0]
-            low_best = 10000000
+            low_best = 1e5
             for i, val in enumerate(fitness_list):
                 low = w_1 * fitness_list[i] + w_2 * fitness_list_2[i]
                 if low < low_best:
@@ -2359,7 +2365,7 @@ class Mutlithreaded_Meta(DifferentialEvolution, SimulatedAnnealing, HarmonySearc
             Temp3 = (Temp1 + Temp2) / (w_1 * Temp2 + w_2 * Temp1)
 
             self.best_energy = slns[0]
-            low_best = 10000000
+            low_best = 1e5
             for i, val in enumerate(fitness_list):
                 low = w_1 * fitness_list[i] + w_2 * fitness_list_2[i]
                 if low < low_best:

{metacountregressor-0.1.315 → metacountregressor-0.1.317}/metacountregressor/solution.py

@@ -123,7 +123,7 @@ class ObjectiveFunction(object):
     """
 
     def __init__(self, x_data, y_data, **kwargs):
-        self.gbl_best = 1000000.0
+        self.gbl_best = 1e5
         self.run_numerical_hessian = kwargs.get('r_nu_hess', False)
         self.run_bootstrap =  kwargs.get('run_bootstrap', False)
         self.linear_regression = kwargs.get('linear_model', False)
@@ -611,6 +611,7 @@ class ObjectiveFunction(object):
 
         modified_fit = self.modify_initial_fit(manual_fit)  # Modify the initial fit based on manual_fit
         self.significant = 1
+        #self.define_selfs_fixed_rdm_cor(modified_fit)
         self.makeRegression(modified_fit)  # Perform regression with the modified fit
 
 
@@ -1571,9 +1572,9 @@ class ObjectiveFunction(object):
             eVy = eVy.ravel()
 
        
-        eVy = np.nan_to_num(eVy, nan=100000, posinf=100000)
-        eVy = np.clip(eVy, None, 1000)
-        mae = np.nan_to_num(MAE(np.squeeze(y), np.squeeze(eVy)), nan=100000, posinf=100000)
+        eVy = np.nan_to_num(eVy, nan=1E5, posinf=1E5)
+        eVy = np.clip(eVy, None, 1E5)
+        mae = np.nan_to_num(MAE(np.squeeze(y), np.squeeze(eVy)), nan=1e5, posinf=1e5)
 
         mspe = np.nan_to_num(MSPE(np.squeeze(y), np.squeeze(eVy)), nan=100000, posinf=100000)
         RMSE = np.sqrt(mspe)
@@ -4797,7 +4798,7 @@ class ObjectiveFunction(object):
                 penalty = self.custom_penalty(betas, penalty)
 
                 if not np.isreal(loglik):
-                    loglik = - 10000000.0
+                    loglik = - 1e5
 
                 output = (-loglik + penalty,)
                 if return_gradient:
@@ -5933,7 +5934,11 @@ class ObjectiveFunction(object):
         """
         X, Xr, XG, XH = mod.get('X'), mod.get('Xr'), mod.get('XG'), mod.get('XH')
         if XG is not None:
-            return np.concatenate((X, XG, Xr, XH), axis=2)
+            if XH is not None:
+                return np.concatenate((X, XG, Xr, XH), axis=2)
+            else:
+                return np.concatenate((X, XG, Xr), axis=2)
+           # return np.concatenate((X, XG, Xr, XH), axis=2)
         elif XH is not None:
             return np.concatenate((X, Xr, XH), axis=2)
         else:
@@ -6287,11 +6292,14 @@ class ObjectiveFunction(object):
         X_test, Xr_test, XG_test, XH_test = (
             mod.get('X_test'), mod.get('Xr_test'), mod.get('XG_test'), mod.get('XH_test')
         )
-        if X_test is None or Xr_test is None:
+        if X_test is None and Xr_test is None:
             return None
 
         if XH_test is not None:
-            return np.concatenate((X_test, Xr_test, XH_test), axis=2)
+            if XG_test is not None:
+                return np.concatenate((X_test, XG_test, Xr_test, XH_test), axis=2)
+            else:
+                return np.concatenate((X_test, Xr_test, XH_test), axis=2)
         elif XG_test is not None:
             return np.concatenate((X_test, XG_test, Xr_test), axis=2)
         else:
@@ -7638,6 +7646,8 @@ class ObjectiveFunction(object):
         alpha_group_rdm = np.isin(select_data, data.get('group_rdm', [])).astype(int).tolist()
         alpha_hetro = np.isin(select_data, [item.split(':')[0] for item in data.get('hetro_in_means', [])]).astype(
             int).tolist()
+        matching_names = [name for name in select_data if name in [item.split(':')[0] for item in data.get('hetro_in_means', [])]]
+        self.hetro_fit = matching_names
         for i in range(len(alpha_rdm)):
 
             if alpha[i]:
@@ -7678,14 +7688,14 @@ class ObjectiveFunction(object):
         # select_data = self._x_data.columns
 
         select_data = self._characteristics_names
-        alpha = np.in1d(select_data, fix) * 1
-        alpha_rdm = np.in1d(select_data, rdm) * 1
+        alpha = np.isin(select_data, fix) * 1
+        alpha_rdm = np.isin(select_data, rdm) * 1
         alpha = alpha.tolist()
         alpha_rdm = alpha_rdm.tolist()
 
-        alpha_cor_rdm = np.in1d(select_data, cor_rdm) * 1
+        alpha_cor_rdm = np.isin(select_data, cor_rdm) * 1
         alpha_cor_rdm = alpha_cor_rdm.tolist()
-        alpha_group_rdm = np.in1d(select_data, group_rdm) * 1
+        alpha_group_rdm = np.isin(select_data, group_rdm) * 1
         alpha_group_rdm = alpha_group_rdm.tolist() #todo will this ever trigger
         return alpha, alpha_rdm, alpha_cor_rdm
 

{metacountregressor-0.1.315 → metacountregressor-0.1.317/metacountregressor.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: metacountregressor
-Version: 0.1.315
+Version: 0.1.317
 Summary: Extensive Testing for Estimation of Data Count Models
 Home-page: https://github.com/zahern/CountDataEstimation
 Author: Zeke Ahern

metacountregressor-0.1.317/version.txt

@@ -0,0 +1 @@
+0.1.317

metacountregressor-0.1.315/version.txt

@@ -1 +0,0 @@
-0.1.315