Merge pull request #7 from BAMeScience/dev

Small hotfixes

fiora/GNN/GNNModules.py

@@ -265,7 +265,7 @@ def load_from_state_dict(cls, PATH: str) -> 'GNNCompiler':
         with open(PARAMS_PATH, 'r') as fp:
             params = json.load(fp)
         model = GNNCompiler(params)
-        model.load_state_dict(torch.load(STATE_PATH))
+        model.load_state_dict(torch.load(STATE_PATH, map_location=torch.serialization.default_restore_location, weights_only=True))
 
         if not isinstance(model, cls):
             raise ValueError(f'file {PATH} contains incorrect model class {type(model)}')
@@ -291,3 +291,22 @@ def save(self, PATH: str, dev: str="cpu") -> None:
 
         #Reset to previous device
         self.to(prev_device)
+
+    # def save(self, PATH: str, dev: str="cpu") -> None:
+
+    #     # Set device to cpu for saving
+    #     prev_device = next(self.parameters()).device
+    #     self.to(dev)
+    #     with open(PATH, 'wb') as f:
+    #         dill.dump(self.to(dev), f)
+
+    #     params_path = '.'.join(PATH.split('.')[:-1]) + '_params.json'
+    #     with open(params_path, 'w') as fp:
+    #         json.dump(self.model_params, fp)
+
+    #     state_dict_path = params_path.replace("_params.json", "_state.pt")
+    #     self.to(dev)
+
+    #     torch.save(self.state_dict(), state_dict_path, _use_new_zipfile_serialization=False)
+
+    #     self.to(prev_device)

notebooks/test_model.ipynb

@@ -9,9 +9,9 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[10:21:18] WARNING: not removing hydrogen atom without neighbors\n",
-      "[10:21:18] WARNING: not removing hydrogen atom without neighbors\n",
-      "[10:21:18] WARNING: not removing hydrogen atom without neighbors\n"
+      "[11:46:34] WARNING: not removing hydrogen atom without neighbors\n",
+      "[11:46:34] WARNING: not removing hydrogen atom without neighbors\n",
+      "[11:46:34] WARNING: not removing hydrogen atom without neighbors\n"
      ]
     },
     {
@@ -68,7 +68,16 @@
    "cell_type": "code",
    "execution_count": 2,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/ynowatzk/miniforge3/envs/fiora/lib/python3.10/site-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly.  To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()\n",
+      "  return self.fget.__get__(instance, owner)()\n"
+     ]
+    }
+   ],
    "source": [
     "depth = 6\n",
     "\n",
@@ -85,16 +94,12 @@
     "from fiora.GNN.GNNModules import GNNCompiler\n",
     "from fiora.MS.SimulationFramework import SimulationFramework\n",
     "\n",
-    "try: \n",
-    "    model = GNNCompiler.load(MODEL_PATH)\n",
+    "\n",
+    "try:\n",
+    "    model = GNNCompiler.load_from_state_dict(MODEL_PATH)\n",
     "except:\n",
-    "    try:\n",
-    "        print(f\"Warning: Failed loading the model {MODEL_PATH}. Fall back: Loading the model from state dictionary.\")\n",
-    "        model = GNNCompiler.load_from_state_dict(MODEL_PATH)\n",
-    "        print(\"Model loaded from state dict without further errors.\")\n",
-    "    except:\n",
-    "        raise NameError(\"Error: Failed loading from state dict.\")\n",
-    "        "
+    "    raise NameError(\"Error: Failed loading from state dict.\")\n",
+    "    "
    ]
   },
   {
@@ -289,13 +294,13 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "/tmp/ipykernel_937446/3302879299.py:24: SettingWithCopyWarning: \n",
+      "/tmp/ipykernel_3323114/3302879299.py:24: SettingWithCopyWarning: \n",
       "A value is trying to be set on a copy of a slice from a DataFrame.\n",
       "Try using .loc[row_indexer,col_indexer] = value instead\n",
       "\n",
       "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
       "  df_train[\"Metabolite\"] = df_train[\"SMILES\"].apply(Metabolite) # TRAIN Metabolites are only tracked for tanimoto distance\n",
-      "/tmp/ipykernel_937446/3302879299.py:25: SettingWithCopyWarning: \n",
+      "/tmp/ipykernel_3323114/3302879299.py:25: SettingWithCopyWarning: \n",
       "A value is trying to be set on a copy of a slice from a DataFrame.\n",
       "Try using .loc[row_indexer,col_indexer] = value instead\n",
       "\n",
@@ -655,7 +660,13 @@
      "output_type": "stream",
      "text": [
       "/home/ynowatzk/repos/fiora/fiora/MS/SimulationFramework.py:170: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.\n",
-      "  df = pd.concat([df, pd.DataFrame(columns=[x + suffix for x in [\"cosine_similarity\", \"kl_div\", \"sim_peaks\", \"spectral_cosine\", \"spectral_sqrt_cosine\", \"spectral_sqrt_cosine_wo_prec\", \"spectral_refl_cosine\", \"spectral_bias\", \"spectral_sqrt_bias\", \"spectral_sqrt_bias_wo_prec\", \"spectral_refl_bias\", \"steins_cosine\", \"steins_bias\", \"RT_pred\", \"RT_dif\", \"CCS_pred\"]])])\n",
+      "  df = pd.concat([df, pd.DataFrame(columns=[x + suffix for x in [\"cosine_similarity\", \"kl_div\", \"sim_peaks\", \"spectral_cosine\", \"spectral_sqrt_cosine\", \"spectral_sqrt_cosine_wo_prec\", \"spectral_refl_cosine\", \"spectral_bias\", \"spectral_sqrt_bias\", \"spectral_sqrt_bias_wo_prec\", \"spectral_refl_bias\", \"steins_cosine\", \"steins_bias\", \"RT_pred\", \"RT_dif\", \"CCS_pred\"]])])\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
       "/home/ynowatzk/repos/fiora/fiora/MS/spectral_scores.py:26: RuntimeWarning: divide by zero encountered in divide\n",
       "  vec = vec / np.linalg.norm(vec)\n",
       "/home/ynowatzk/repos/fiora/fiora/MS/spectral_scores.py:27: RuntimeWarning: divide by zero encountered in divide\n",
@@ -1650,7 +1661,7 @@
        "      <td>0.717736</td>\n",
        "      <td>0.780847</td>\n",
        "      <td>0.273686</td>\n",
-       "      <td>0.307230</td>\n",
+       "      <td>0.307231</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
@@ -1687,7 +1698,7 @@
        "2       ICEBERG  0.716473       NaN  0.575536       NaN  0.709359       NaN   \n",
        "\n",
        "   CASMI22+  CASMI22-  \n",
-       "0  0.273686  0.307230  \n",
+       "0  0.273686  0.307231  \n",
        "1  0.376012  0.292884  \n",
        "2  0.358571  0.000000  "
       ]
@@ -2011,17 +2022,17 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "cos: 0.7399321531311543 0.5158647138450783\n",
-      "cos: 0.7399321531311543 0.3008378811624405\n",
-      "cos: 0.7399321531311543 0.3014740423593587\n",
-      "cos: 0.5824634669347201 0.5158647138450783\n",
-      "cos: 0.5824634669347201 0.3008378811624405\n",
-      "cos: 0.5824634669347201 0.3014740423593587\n",
-      "cos: 0.5824634893228796 0.5158647138450783\n",
-      "cos: 0.5824634893228796 0.3008378811624405\n",
-      "cos: 0.5824634893228796 0.3014740423593587\n",
-      "cos: 0.5780898248582669 0.4830928581051953\n",
-      "cos: 0.49608554398224664 0.4830928581051953\n"
+      "cos: 0.7399321707609304 0.5158647125171295\n",
+      "cos: 0.7399321707609304 0.3008378766012895\n",
+      "cos: 0.7399321707609304 0.30147404234591146\n",
+      "cos: 0.5824634973655437 0.5158647125171295\n",
+      "cos: 0.5824634973655437 0.3008378766012895\n",
+      "cos: 0.5824634973655437 0.30147404234591146\n",
+      "cos: 0.5824634973655437 0.5158647125171295\n",
+      "cos: 0.5824634973655437 0.3008378766012895\n",
+      "cos: 0.5824634973655437 0.30147404234591146\n",
+      "cos: 0.5780898227574808 0.48309285488696774\n",
+      "cos: 0.4960855687939471 0.48309285488696774\n"
      ]
     }
    ],

scripts/fiora-predict

@@ -1,10 +1,13 @@
 #! /usr/bin/env python
 import pandas as pd
 import os
+import warnings
 from rdkit import RDLogger
 RDLogger.DisableLog("rdApp.*")
+warnings.filterwarnings("ignore", category=UserWarning, message="TypedStorage is deprecated")
 
 import argparse
+import importlib.resources
 import fiora.IO.mgfWriter as mgfWriter
 import fiora.IO.mspWriter as mspWriter
 
@@ -15,7 +18,6 @@ from fiora.GNN.AtomFeatureEncoder import AtomFeatureEncoder
 from fiora.GNN.BondFeatureEncoder import BondFeatureEncoder
 from fiora.GNN.SetupFeatureEncoder import SetupFeatureEncoder
 
-
 def parse_args() -> argparse.Namespace:
     parser = argparse.ArgumentParser(prog='fiora-predict',
                     description='Fiora is an in silico fragmentation framework, which predicts peaks and simulates tandem mass spectra including features such as retention time and collision cross sections. Use this script for spectrum predictions with a (pre-)trained model.',
@@ -28,7 +30,7 @@ def parse_args() -> argparse.Namespace:
 
 
     parser.add_argument('--rt', action=argparse.BooleanOptionalAction, help="predict retention time", default=False)
-    parser.add_argument('--ccs', action=argparse.BooleanOptionalAction, help="predict collison cross section", default=True)
+    parser.add_argument('--ccs', action=argparse.BooleanOptionalAction, help="predict collison cross section", default=False)
     parser.add_argument("--annotation", action=argparse.BooleanOptionalAction, help="annotate predicted peaks with SMILES strings", default=False)
     args = parser.parse_args()
 
@@ -102,19 +104,17 @@ def main():
     print(f'Running fiora prediction with the following parameters: {args}\n')
 
     # Load model
-    if args.model == "default": 
-        args.model = os.path.join(os.path.dirname(os.path.abspath(__file__)), "../models/fiora_OS_v0.1.0.pt")
+    if args.model == "default":
+        with importlib.resources.path('models', 'fiora_OS_v0.1.0.pt') as model_path:
+            args.model = str(model_path)
+        #os.path.join(os.path.dirname(os.path.abspath(__file__)), "../models/fiora_OS_v0.1.0.pt")
 
-    try: 
-        model = GNNCompiler.load(args.model)
-    except:
-        try:
-            print(f"Warning: Failed loading the model {args.model}. Fallback: Loading the model from state dictionary.")
-            model = GNNCompiler.load_from_state_dict(args.model)
-            print("Model loaded from state dict without further errors.")
-        except Exception as e:
-            print(f"Error: Failed loading from state dict. Caused by: {e}.")
-            exit(1)
+    print(args.model)
+    try:
+        model = GNNCompiler.load_from_state_dict(args.model)
+    except Exception as e:
+        print(f"Error: Failed loading from model from state dict. Caused by: {e}.")
+        exit(1)
 
     print_model_messages(model.model_params)
     args = update_args_with_model_params(args, model.model_params)

setup.py

@@ -1,6 +1,6 @@
 from setuptools import setup
 
-setup(name='Fiora',
+setup(name='fiora',
       version='0.0.1',
       long_description='file: README.md',
       author='Yannek Nowatzky',
@@ -10,14 +10,13 @@
           'Operating System :: OS Independent',
           'Programming Language :: Python :: 3',
       ],
-      # entry_points={
-      #   "console_scripts": [
-      #       "fiora-predict = scripts.predict:main",
-      #   ],
-      #   },
+
       scripts=["scripts/fiora-predict"],
+      packages=['fiora', 'fiora.GNN', 'fiora.IO',  'fiora.MOL', 'fiora.MS', 'fiora.visualization', 'models'],
       include_package_data=True,
-      packages=['fiora', 'fiora.GNN', 'fiora.IO',  'fiora.MOL', 'fiora.MS', 'fiora.visualization'],
+      package_data={
+        'models': ['fiora_OS_v0.1.0.pt', 'fiora_OS_v0.1.0_state.pt', 'fiora_OS_v0.1.0_params.json'],
+      },
       install_requires=['numpy', 'seaborn', 'torch', 'torch_geometric', 'dill', 'rdkit', 'treelib', 'spectrum_utils', 'setuptools>=24.2.0'],
       python_requires='>=3.10.8',
       # Developers may also want to install: jupyter torchmetrics umap umap-learn pytest