Merge pull request #59 from instadeepai/58-instanovo-100-code-tests

InstaNovo 1.0.0 code tests

.github/workflows/pytest-multiversion.yml

@@ -15,7 +15,7 @@ jobs:
       fail-fast: false
       matrix:
         python-version: ["3.10", "3.11"] # "3.12"
-        os: [ubuntu-latest, windows-latest] # macos-latest
+        os: [ubuntu-latest] #windows-latest # macos-latest
     runs-on: ${{ matrix.os }}
 
     steps:
@@ -36,6 +36,7 @@ jobs:
           pre-commit run --all-files -c .pre-commit-config.yaml
       - name: Test with pytest
         run: |
+          python -m instanovo.scripts.get_zenodo_record
           pytest -v --alluredir=allure_results  --cov-report=html --cov  --cov-config=.coveragerc --random-order
           coverage report -m
       - name: Test notebooks on Linux

.gitignore

@@ -171,7 +171,6 @@ docs/reference
 coverage/
 checkpoints/
 data/
-docs/
 docs_public/
 logs/
 mlruns/

docs/LICENSE.md

@@ -0,0 +1 @@
+--8<-- "LICENSE.md"

docs/gen_ref_nav.py

@@ -0,0 +1,68 @@
+"""Generate the code reference pages and navigation."""
+
+# This script is used by the mkdocs-gen-files plugin (https://oprypin.github.io/mkdocs-gen-files/)
+# for MkDocs (https://www.mkdocs.org/). It creates for each module in the code a stub page
+# and it creates a "docs/reference/SUMMARY.md" page which contains a Table of Contents with links to
+# all the stub pages. When MkDocs runs, it will populate the stub pages with the documentation
+# pulled from the docstrings
+from __future__ import annotations
+
+from pathlib import Path
+
+import mkdocs_gen_files
+
+# Folders for which we don't want to create code documentation but which can contain *.py files
+IGNORE_DIRS = ("build", "docs_public", "docs", "tests", "scripts", "utils", ".venv")
+
+
+def is_ignored_directory(module_path: Path) -> bool:
+    """Check if the module path is within any ignored directory."""
+    return any(part in IGNORE_DIRS for part in module_path.parts)
+
+
+def is_ignored_file(module_path: Path) -> bool:
+    """Check if the file is a test file or ignored file."""
+    return module_path.parts[-1].endswith("_test") or module_path.parts[-1] in (
+        "mlflow_auth",
+        "types",
+        "constants",
+    )
+
+
+def process_python_files(source_directory: str, module_name: str) -> None:
+    """Generate documentation paths for Python files in the source directory."""
+    nav = mkdocs_gen_files.Nav()
+
+    for python_file in sorted(Path(source_directory).rglob("*.py")):
+        relative_module_path = python_file.relative_to(source_directory).with_suffix("")
+
+        if not is_ignored_directory(relative_module_path) and not is_ignored_file(
+            relative_module_path
+        ):
+            doc_path = python_file.relative_to(
+                source_directory, module_name
+            ).with_suffix(".md")
+            full_doc_path = Path("reference", doc_path)
+
+            parts = tuple(relative_module_path.parts)
+
+            if parts[-1] == "__init__":
+                parts = parts[:-1]
+                doc_path = doc_path.with_name("index.md")
+                full_doc_path = full_doc_path.with_name("index.md")
+            elif parts[-1] == "__main__":
+                continue
+
+            nav[parts] = doc_path.as_posix()
+
+            with mkdocs_gen_files.open(full_doc_path, "w") as fd:
+                ident = ".".join(parts)
+                fd.write(f"::: {ident}")
+
+            mkdocs_gen_files.set_edit_path(full_doc_path, ".." / python_file)
+
+    with mkdocs_gen_files.open("reference/SUMMARY.md", "w") as nav_file:
+        nav_file.writelines(nav.build_literate_nav())
+
+
+process_python_files(source_directory=".", module_name="instanovo")

docs/index.md

@@ -0,0 +1 @@
+--8<-- "README.md"

instanovo/configs/model/instanovo_8M.yaml

@@ -0,0 +1,10 @@
+defaults:
+  - instanovo_base
+  - _self_
+
+# Model parameters
+dim_model: 320
+n_head: 20
+dim_feedforward: 1280
+n_layers: 6
+dropout: 0.

tests/conftest.py

@@ -0,0 +1,147 @@
+from __future__ import annotations
+
+import os
+import random
+import sys
+from typing import Any
+
+import numpy as np
+import pandas as pd
+import pytest
+import pytorch_lightning as ptl
+import torch
+from hydra import compose, initialize
+from omegaconf import DictConfig, open_dict
+
+from instanovo.inference.knapsack_beam_search import KnapsackBeamSearchDecoder
+from instanovo.transformer.model import InstaNovo
+from instanovo.transformer.predict import _setup_knapsack
+
+
+# Add the root directory to the PYTHONPATH
+# This allows pytest to find the modules for testing
+
+root_dir = os.path.dirname(os.path.dirname(__file__))
+sys.path.append(root_dir)
+
+
+def reset_seed(seed: int = 42) -> None:
+    """Function to reset seeds."""
+    torch.manual_seed(seed)
+    if torch.cuda.is_available():
+        torch.cuda.manual_seed_all(seed)
+    np.random.seed(seed)
+    random.seed(seed)
+    ptl.seed_everything(seed)
+
+
+@pytest.fixture()
+def _reset_seed() -> None:
+    """A pytest fixture to reset the seeds at the start of relevant tests."""
+    reset_seed()
+
+
+@pytest.fixture(scope="session")
+def checkpoints_dir() -> str:
+    """A pytest fixture to create and provide the absolute path of a 'checkpoints' directory.
+
+    Ensures the directory exists for storing checkpoint files during the test session.
+    """
+    checkpoints_dir = "checkpoints"
+    os.makedirs(checkpoints_dir, exist_ok=True)
+    return os.path.abspath(checkpoints_dir)
+
+
+@pytest.fixture(scope="session")
+def instanovo_config() -> DictConfig:
+    """A pytest fixture to read in a Hydra config for the Instanovo model unit and integration tests."""
+    with initialize(version_base=None, config_path="../instanovo/configs"):
+        cfg = compose(config_name="instanovo_unit_test")
+
+    sub_configs_list = ["model", "dataset", "residues"]
+    for sub_name in sub_configs_list:
+        if sub_name in cfg:
+            with open_dict(cfg):
+                temp = cfg[sub_name]
+                del cfg[sub_name]
+                cfg.update(temp)
+
+    return cfg
+
+
+@pytest.fixture(scope="session")
+def instanovo_inference_config() -> DictConfig:
+    """A pytest fixture to read in a Hydra config for inference of the Instanovo model unit and integration tests."""
+    with initialize(version_base=None, config_path="../instanovo/configs/inference"):
+        cfg = compose(config_name="unit_test")
+
+    return cfg
+
+
+@pytest.fixture(scope="session")
+def dir_paths() -> tuple[str, str]:
+    """A pytest fixture that returns the root and data directories for the unit and integration tests."""
+    root_dir = "./tests/instanovo_test_resources"
+    data_dir = os.path.join(root_dir, "example_data")
+    return root_dir, data_dir
+
+
+@pytest.fixture(scope="session")
+def instanovo_checkpoint(dir_paths: tuple[str, str]) -> str:
+    """A pytest fixture that returns the InstaNovo model checkpoint used for unit and integration tests."""
+    root_dir, _ = dir_paths
+    return os.path.join(root_dir, "model.ckpt")
+
+
+@pytest.fixture(scope="session")
+def instanovo_model(
+    instanovo_checkpoint: str,
+) -> tuple[Any, Any]:
+    """A pytest fixture that returns the InstaNovo model and config used for unit and integration tests."""
+    model, config = InstaNovo.load(path=instanovo_checkpoint)
+    return model, config
+
+
+@pytest.fixture(scope="session")
+def residue_set(instanovo_model: tuple[Any, Any]) -> Any:
+    """A pytest fixture to return the model's residue set used for unit and integration tests."""
+    model, _ = instanovo_model
+    return model.residue_set
+
+
+@pytest.fixture(scope="session")
+def instanovo_output(dir_paths: tuple[str, str]) -> pd.DataFrame:
+    """A pytest fixture to load the pre-computed InstaNovo model predictions for unit and integration tests."""
+    root_dir, _ = dir_paths
+    return pd.read_csv(os.path.join(root_dir, "predictions.csv"))
+
+
+@pytest.fixture(scope="session")
+def knapsack_dir(dir_paths: tuple[str, str]) -> str:
+    """A pytest fixture to create and provide the absolute path of a 'knapsack' directory within the checkpoints directory for storing test artifacts."""
+    root_dir, _ = dir_paths
+    knapsack_dir = os.path.join(root_dir, "example_knapsack")
+    return os.path.abspath(knapsack_dir)
+
+
+@pytest.fixture(scope="session")
+def setup_knapsack_decoder(
+    instanovo_model: tuple[Any, Any], knapsack_dir: str
+) -> KnapsackBeamSearchDecoder:
+    """A pytest fixture to create a Knapsack object."""
+    model, _ = instanovo_model
+
+    if os.path.exists(knapsack_dir):
+        decoder = KnapsackBeamSearchDecoder.from_file(model=model, path=knapsack_dir)
+        print("Loaded knapsack decoder.")
+
+    else:
+        knapsack = _setup_knapsack(model)
+
+        knapsack.save(path=knapsack_dir)
+        print("Created and saved knapsack.")
+
+        decoder = KnapsackBeamSearchDecoder(model, knapsack)
+        print("Loaded knapsack decoder.")
+
+    return decoder