This commit is related to issue: https://github.com/EMI-Group/tensorneat/issues/11

1. Add origin_node and origin_conn.
2. Change the behavior of crossover and mutation. Now, TensorNEAT will use all fix_attrs(include historical marker if it has one) as identifier for gene in crossover and distance calculation.
3. Other slightly change.
4. Add two related examples: xor_origin and hopper_origin
5. Add related test file.

examples/brax/hopper_origin.py

@@ -0,0 +1,49 @@
+from tensorneat.pipeline import Pipeline
+from tensorneat.algorithm.neat import NEAT
+from tensorneat.genome import DefaultGenome, OriginNode, OriginConn
+
+from tensorneat.problem.rl import BraxEnv
+from tensorneat.common import ACT, AGG
+
+"""
+Solving Hopper with OriginGene
+See https://github.com/EMI-Group/tensorneat/issues/11
+"""
+
+if __name__ == "__main__":
+    pipeline = Pipeline(
+        algorithm=NEAT(
+            pop_size=1000,
+            species_size=20,
+            survival_threshold=0.1,
+            compatibility_threshold=1.0,
+            genome=DefaultGenome(
+                num_inputs=11,
+                num_outputs=3,
+                init_hidden_layers=(),
+                # origin node gene, which use the same crossover behavior to the origin NEAT paper
+                node_gene=OriginNode(
+                    activation_options=ACT.tanh,
+                    aggregation_options=AGG.sum,
+                    response_lower_bound = 1,
+                    response_upper_bound= 1,  # fix response to 1
+                ),
+                # use origin connection, which using historical marker
+                conn_gene=OriginConn(),  
+                output_transform=ACT.tanh,
+            ),
+        ),
+        problem=BraxEnv(
+            env_name="hopper",
+            max_step=1000,
+        ),
+        seed=42,
+        generation_limit=100,
+        fitness_target=5000,
+    )
+
+    # initialize state
+    state = pipeline.setup()
+    # print(state)
+    # run until terminate
+    state, best = pipeline.auto_run(state)

examples/func_fit/xor.py

@@ -30,7 +30,8 @@ pipeline.show(state, best)
 
 # visualize the best individual
 network = algorithm.genome.network_dict(state, *best)
-algorithm.genome.visualize(network, save_path="./imgs/xor_network.svg")
+print(algorithm.genome.repr(state, *best))
+# algorithm.genome.visualize(network, save_path="./imgs/xor_network.svg")
 
 # transform the best individual to latex formula
 from tensorneat.common.sympy_tools import to_latex_code, to_python_code

examples/func_fit/xor_origin.py

@@ -0,0 +1,55 @@
+from tensorneat.pipeline import Pipeline
+from tensorneat import algorithm, genome, problem
+from tensorneat.genome import OriginNode, OriginConn
+from tensorneat.common import ACT
+
+"""
+Solving XOR-3d problem with OriginGene
+See https://github.com/EMI-Group/tensorneat/issues/11
+"""
+
+algorithm = algorithm.NEAT(
+    pop_size=10000,
+    species_size=20,
+    survival_threshold=0.01,
+    genome=genome.DefaultGenome(
+        node_gene=OriginNode(),
+        conn_gene=OriginConn(),
+        num_inputs=3,
+        num_outputs=1,
+        max_nodes=7,
+        output_transform=ACT.sigmoid,
+    ),
+)
+problem = problem.XOR3d()
+
+pipeline = Pipeline(
+    algorithm,
+    problem,
+    generation_limit=200,
+    fitness_target=-1e-6,
+    seed=42,
+)
+state = pipeline.setup()
+# run until terminate
+state, best = pipeline.auto_run(state)
+# show result
+pipeline.show(state, best)
+
+# visualize the best individual
+network = algorithm.genome.network_dict(state, *best)
+print(algorithm.genome.repr(state, *best))
+# algorithm.genome.visualize(network, save_path="./imgs/xor_network.svg")
+
+# transform the best individual to latex formula
+from tensorneat.common.sympy_tools import to_latex_code, to_python_code
+
+sympy_res = algorithm.genome.sympy_func(
+    state, network, sympy_output_transform=ACT.obtain_sympy(ACT.sigmoid)
+)
+latex_code = to_latex_code(*sympy_res)
+print(latex_code)
+
+# transform the best individual to python code
+python_code = to_python_code(*sympy_res)
+print(python_code)