[Feat] automatic testing

.github/workflows/tests.yml

@@ -0,0 +1,38 @@
+name: Tests
+on: [push, pull_request]
+
+jobs:
+  build:
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: true
+      max-parallel: 15
+      matrix:
+        os: [ubuntu-latest, macos-latest, windows-latest]
+        python-version: ['3.11']
+    defaults:
+      run:
+        shell: bash
+    steps:
+    - name: Check out repository
+      uses: actions/checkout@v3
+
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v4
+      with:
+        python-version: ${{ matrix.python-version }}
+
+    - name: Load cached venv
+      id: cached-pip-wheels
+      uses: actions/cache@v3
+      with:
+        path: ~/.cache
+        key: venv-${{ runner.os }}-${{ steps.setup-python.outputs.python-version }}-${{ hashFiles('**/poetry.lock') }}
+
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install -e ".[all]"
+
+    - name: Run pytest
+      run: pytest tests/*.py

tests/env.py

@@ -0,0 +1,46 @@
+import pytest
+
+from routefinder.envs.mtvrp import MTVRPEnv, MTVRPGenerator
+from routefinder.models import RouteFinderPolicy
+from routefinder.utils import greedy_policy, rollout
+
+
+@pytest.mark.parametrize(
+    "variant_preset",
+    [
+        "all",
+        "single_feat",
+        "single_feat_otw",
+        "cvrp",
+        "ovrp",
+        "vrpb",
+        "vrpl",
+        "vrptw",
+        "ovrptw",
+        "ovrpb",
+        "ovrpl",
+        "vrpbl",
+        "vrpbtw",
+        "vrpltw",
+        "ovrpbl",
+        "ovrpbtw",
+        "ovrpltw",
+        "vrpbltw",
+        "ovrpbltw",
+    ],
+)
+def test_env(variant_preset):
+    # Sample all variants in the same batch (Mixed-Batch Training)
+    generator = MTVRPGenerator(num_loc=10, variant_preset=variant_preset)
+    env = MTVRPEnv(generator, check_solution=True)
+    td_data = env.generator(3)
+    td_test = env.reset(td_data)
+    actions = rollout(env, td_test.clone(), greedy_policy)
+    rewards_nearest_neighbor = env.get_reward(td_test, actions)
+    assert rewards_nearest_neighbor.shape == (3,)
+
+    policy = RouteFinderPolicy()
+    out = policy(
+        td_test.clone(), env, phase="test", decode_type="greedy", return_actions=True
+    )
+    assert out["reward"].shape == (3,)

tests/policy.py

@@ -0,0 +1,41 @@
+import pytest
+
+from routefinder.envs.mtvrp import MTVRPEnv, MTVRPGenerator
+from routefinder.models import RouteFinderPolicy
+
+
+@pytest.mark.parametrize(
+    "variant_preset",
+    [
+        "all",
+        "single_feat",
+        "single_feat_otw",
+        "cvrp",
+        "ovrp",
+        "vrpb",
+        "vrpl",
+        "vrptw",
+        "ovrptw",
+        "ovrpb",
+        "ovrpl",
+        "vrpbl",
+        "vrpbtw",
+        "vrpltw",
+        "ovrpbl",
+        "ovrpbtw",
+        "ovrpltw",
+        "vrpbltw",
+        "ovrpbltw",
+    ],
+)
+def test_policy(variant_preset):
+    # Sample all variants in the same batch (Mixed-Batch Training)
+    generator = MTVRPGenerator(num_loc=10, variant_preset=variant_preset)
+    env = MTVRPEnv(generator, check_solution=True)
+    td_data = env.generator(3)
+    td_test = env.reset(td_data)
+    policy = RouteFinderPolicy()
+    out = policy(
+        td_test.clone(), env, phase="test", decode_type="greedy", return_actions=True
+    )
+    assert out["reward"].shape == (3,)

tests/training.py

@@ -0,0 +1,26 @@
+from rl4co.utils.trainer import RL4COTrainer
+
+from routefinder.envs.mtvrp import MTVRPEnv
+from routefinder.models import RouteFinderBase, RouteFinderPolicy
+
+
+def test_training():
+    env = MTVRPEnv(generator_params={"num_loc": 10, "variant_preset": "all"})
+    policy = RouteFinderPolicy()
+    model = RouteFinderBase(
+        env,
+        policy,
+        batch_size=3,
+        train_data_size=3,
+        val_data_size=3,
+        test_data_size=3,
+        optimizer_kwargs={"lr": 3e-4, "weight_decay": 1e-6},
+    )
+    trainer = RL4COTrainer(
+        max_epochs=1,
+        gradient_clip_val=None,
+        devices=1,
+        accelerator="auto",
+    )
+    trainer.fit(model)
+    trainer.test(model)