Make new CNN model as default (Remove old CNN model)

ProjectTorreyPines · Oct 31, 2024 · 5cf211e · 5cf211e
1 parent f3b6865
commit 5cf211e
Show file tree

Hide file tree

Showing 3 changed files with 13 additions and 41 deletions.
diff --git a/data/CNN_Model.onnx b/data/CNN_Model.onnx
diff --git a/data/CNN_Model_new.onnx b/data/CNN_Model_new.onnx
diff --git a/src/TroyonBetaNN.jl b/src/TroyonBetaNN.jl
@@ -415,47 +415,19 @@ end
 
 function _set_CNN_input_neurons_from_sampled_points(TD::Troyon_Data, eqt::IMAS.equilibrium__time_slice)
 
-    if (basename(TD.CNN.filePath) == "CNN_Model.onnx")
-        Xb = _convert_RZ_samples_into_19_normalized_neurons(TD, eqt)
-        Xp = TD.sampPoints.pressure[2:end] / TD.sampPoints.pressure[1]
-        Xq = TD.sampPoints.q
-
-        input_1 = Float32.(reshape(Xb, 1, 19, 1)) # Boundary input
-        input_2 = Float32.(reshape(Xp, 1, 11, 1)) # Pressure
-        input_3 = Float32.(reshape(Xq, 1, 12, 1)) # Safety factor
-
-        input_4 = Float32.(reshape([eqt.global_quantities.li_3], 1, 1)) # internal inductance
-
-        # Calculate PPF (Pressure Peaking Factor)
-        PPF = eqt.profiles_1d.pressure[1] / Take_1D_average_over_volume(eqt, eqt.profiles_1d.pressure)
-        input_5 = Float32.(reshape([PPF], 1, 1))
-
-        #  TODO: check what exactly input_4 and input_5 are
-        # Here, we assume that input_4 is "PPF" and input_5 is "li", which looks more reasonable
-        # Swap input_4 and input_5
-        # input_4, input_5 = input_5, input_4
-    elseif (basename(TD.CNN.filePath) == "CNN_Model_new.onnx")
-        println("using new CNN model")
-        Xb = _convert_RZ_samples_into_19_normalized_neurons(TD, eqt)
-        Xp = TD.sampPoints.pressure[2:end] / TD.sampPoints.pressure[1]
-        Xq = TD.sampPoints.q
-
-        input_1 = Float32.(reshape(Xb, 1, 19, 1)) # Boundary input
-        input_2 = Float32.(reshape(Xq, 1, 12, 1)) # Safety factor
-        input_3 = Float32.(reshape(Xp, 1, 11, 1)) # Pressure
-
-        input_4 = Float32.(reshape([eqt.global_quantities.li_3], 1, 1)) # internal inductance
-
-        # Calculate PPF (Pressure Peaking Factor)
-        PPF = eqt.profiles_1d.pressure[1] / Take_1D_average_over_volume(eqt, eqt.profiles_1d.pressure)
-        input_5 = Float32.(reshape([PPF], 1, 1))
-
-
-        #  TODO: check what exactly input_4 and input_5 are
-        # Here, we assume that input_4 is "PPF" and input_5 is "li", which looks more reasonable
-        # Swap input_4 and input_5
-        # input_4, input_5 = input_5, input_4
-    end
+    Xb = _convert_RZ_samples_into_19_normalized_neurons(TD, eqt)
+    Xp = TD.sampPoints.pressure[2:end] / TD.sampPoints.pressure[1]
+    Xq = TD.sampPoints.q
+
+    input_1 = Float32.(reshape(Xb, 1, 19, 1)) # Boundary input
+    input_2 = Float32.(reshape(Xq, 1, 12, 1)) # Safety factor
+    input_3 = Float32.(reshape(Xp, 1, 11, 1)) # Pressure
+
+    input_4 = Float32.(reshape([eqt.global_quantities.li_3], 1, 1)) # internal inductance
+
+    # Calculate PPF (Pressure Peaking Factor)
+    PPF = eqt.profiles_1d.pressure[1] / Take_1D_average_over_volume(eqt, eqt.profiles_1d.pressure)
+    input_5 = Float32.(reshape([PPF], 1, 1))
 
     return TD.CNN.input = Dict("input_1" => input_1, "input_2" => input_2, "input_3" => input_3, "input_4" => input_4, "input_5" => input_5)
 end