Postprocessing

xppbe/Post/Post_Template.ipynb

@@ -55,7 +55,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "Iter = 20000\n",
+    "Iter = 'last'\n",
     "simulation.load_model_for_Post(Iter, save=False)\n",
     "Post = simulation.Post"
    ]

xppbe/Post/Results_utils.py

@@ -3,12 +3,16 @@
 import pandas as pd
 
 
-def find_largest_iteration(folder_path):
+def get_max_iteration(folder_path):
     subdirectories = [d for d in os.listdir(folder_path) if os.path.isdir(os.path.join(folder_path, d))]
     iteration_numbers = [int(d.split('_')[1]) for d in subdirectories if d.startswith('iter_')]
     if not iteration_numbers:
         return None
-    largest_iteration_folder = os.path.join(folder_path, f'iter_{max(iteration_numbers)}')
+    return max(iteration_numbers)
+
+def find_largest_iteration(folder_path):
+    Iter = get_max_iteration(folder_path)
+    largest_iteration_folder = os.path.join(folder_path, f'iter_{Iter}')
     return largest_iteration_folder
 
 def find_loss_csv(folder_path):

xppbe/Simulation.py

@@ -175,14 +175,18 @@ def postprocessing(self, run_all=False, jupyter=False, plot_mesh=False, known_me
         return self.Post
 
 
-    def load_model_for_Post(self,Iter,save=False):
+    def load_model_for_Post(self,Iter='last',save=False):
 
         if self.num_networks == 2:
             from xppbe.NN.NeuralNet import PINN_2Dom_NeuralNet as NeuralNet
         elif self.num_networks == 1:
             from xppbe.NN.NeuralNet import PINN_1Dom_NeuralNet as NeuralNet  
 
         self.PINN_solver.results_path = self.results_path
+        if Iter=='last':
+            from xppbe.Post.Results_utils import get_max_iteration
+            Iter = get_max_iteration(os.path.join(self.results_path,'iterations'))
+
         self.PINN_solver.load_NeuralNet(NeuralNet,self.results_path,f'iter_{Iter}',Iter)
         self.PINN_solver.N_iters = self.N_iters