model: Consolidated self.location as a property of baseclass

- model: pytorch: Updated to use self.location property
- model: spacy: Updated to use self.location property
- model: tensorflow_hub: Updated to use self.location property

CHANGELOG.md

@@ -36,7 +36,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Dev CMD to remove unused imports, `$ dffml service dev lint imports`
 - Helper for creating a blank generic Python package
   `$ dffml service dev create blank mypackage`
-- Added `is_trained` flag to all models
+- `is_trained` flag to all models
+- Dynamic `location` property to `Model` baseclass.
 ### Changed
 - Calls to hashlib now go through helper functions
 - Build docs using `dffml service dev docs`

dffml/model/model.py

@@ -116,8 +116,7 @@ async def __aenter__(self):
                     in ["zip", "tar"],
                 ]
             ):
-                temp_dir = self._get_directory()
-                self.location = temp_dir
+                self.create_temp_directory()
             else:
                 self._make_config_location()
 
@@ -196,10 +195,9 @@ async def _run_operation(self, input_path, output_path, dataflow):
         async for _, _ in run(dataflow):
             pass
 
-    def _get_directory(self) -> pathlib.Path:
+    def create_temp_directory(self):
         if not hasattr(self, "temp_dir"):
             self.temp_dir = pathlib.Path(mkdtemp())
-        return self.temp_dir
 
     def _make_config_location(self):
         """
@@ -212,6 +210,14 @@ def _make_config_location(self):
             if not location.is_dir():
                 location.mkdir(mode=MODE_BITS_SECURE, parents=True)
 
+    @property
+    def location(self):
+        return (
+            self.config.location
+            if not hasattr(self, "temp_dir")
+            else self.temp_dir
+        )
+
 
 class SimpleModelNoContext:
     """
@@ -298,12 +304,7 @@ def disk_path(self, extention: Optional[str] = None):
         if "features" in exported:
             exported["features"] = dict(sorted(exported["features"].items()))
         # Hash the exported config
-        return pathlib.Path(
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir,
-            "Model",
-        )
+        return pathlib.Path(self.location, "Model",)
 
     def applicable_features(self, features):
         usable = []

model/autosklearn/dffml_model_autosklearn/autosklearn_base.py

@@ -20,12 +20,7 @@ def __init__(self, config):
 
     async def __aenter__(self) -> "AutoSklearnModel":
         await super().__aenter__()
-        self.path = self.filepath(
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir,
-            "trained_model.sav",
-        )
+        self.path = self.filepath(self.location, "trained_model.sav",)
         self.load_model()
         return self
 

model/pytorch/dffml_model_pytorch/pytorch_base.py

@@ -362,17 +362,9 @@ def createModel(self):
             "Can't use createModel method from PyTorchModel"
         )
 
-    @property
-    def base_path(self):
-        return (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-
     @property
     def model_path(self):
-        return self.base_path / "model.pt"
+        return self.location / "model.pt"
 
     def _classifications(self, cids):
         """

model/scikit/dffml_model_scikit/scikit_base.py

@@ -58,20 +58,12 @@ def __init__(self, config) -> None:
         self.clf = None
         self.joblib = importlib.import_module("joblib")
 
-    @property
-    def _filepath(self):
-        return (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-
     async def __aenter__(self) -> "Scikit":
         await super().__aenter__()
-        self.saved_filepath = self._filepath / "Scikit.json"
+        self.saved_filepath = self.location / "Scikit.json"
         if self.saved_filepath.is_file():
             self.saved = json.loads(self.saved_filepath.read_text())
-        self.clf_path = self._filepath / "ScikitFeatures.joblib"
+        self.clf_path = self.location / "ScikitFeatures.joblib"
         if self.clf_path.is_file():
             self.clf = self.joblib.load(str(self.clf_path))
             self.is_trained = True
@@ -87,10 +79,10 @@ async def __aexit__(self, exc_type, exc_value, traceback):
 class ScikitUnsprvised(Scikit):
     async def __aenter__(self) -> "ScikitUnsprvised":
         await super().__aenter__()
-        self.saved_filepath = self._filepath / "Scikit.json"
+        self.saved_filepath = self.location / "Scikit.json"
         if self.saved_filepath.is_file():
             self.saved = json.loads(self.saved_filepath.read_text())
-        self.clf_path = self._filepath / "ScikitUnsupervised.json"
+        self.clf_path = self.location / "ScikitUnsupervised.json"
         if self.clf_path.is_file():
             self.clf = self.joblib.load(str(self.clf_path))
             self.is_trained = True

model/spacy/dffml_model_spacy/ner/ner_model.py

@@ -295,12 +295,7 @@ def __init__(self, config):
 
     async def __aenter__(self):
         await super().__aenter__()
-        self.path = (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-        self.model_path = self.path / "ner"
+        self.model_path = self.location / "ner"
         if self.model_path.exists():
             self.nlp = spacy.load(self.model_path)
             self.logger.debug("Loaded model from disk.")

model/tensorflow/dffml_model_tensorflow/tf_base.py

@@ -135,17 +135,9 @@ def _applicable_features(self):
             if name in self.feature_columns
         ]
 
-    @property
-    def base_path(self):
-        return (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-
     @property
     def model_path(self):
-        return self.base_path / "DNNModel"
+        return self.location / "DNNModel"
 
     async def __aenter__(self):
         await super().__aenter__()

model/tensorflow_hub/dffml_model_tensorflow_hub/text_classifier.py

@@ -350,14 +350,6 @@ def __init__(self, config):
     def model(self):
         return self._model
 
-    @property
-    def base_path(self):
-        return (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-
     @property
     def model_folder_path(self):
         _to_hash = self.features + [
@@ -366,7 +358,7 @@ def model_folder_path(self):
             self.config.model_path,
         ]
         model_name = secure_hash("".join(_to_hash), algorithm="sha384")
-        model_folder_path = self.base_path / model_name
+        model_folder_path = self.location / model_name
         if not model_folder_path.exists():
             model_folder_path.mkdir(parents=True, exist_ok=True)
         return model_folder_path

model/vowpalWabbit/dffml_model_vowpalWabbit/vw_base.py

@@ -324,14 +324,6 @@ def __init__(self, config) -> None:
         self.clf = None
         self.is_trained = self.model_filename.exists()
 
-    @property
-    def _base_path(self):
-        return (
-            self.config.location
-            if not hasattr(self, "temp_dir")
-            else self.temp_dir
-        )
-
     def applicable_features(self, features):
         usable = []
         for feature in features:
@@ -364,7 +356,7 @@ def _feature_predict_hash(self):
     @property
     def model_filename(self):
         model_name = self._features_hash + ".vw"
-        return self._base_path / model_name
+        return self.location / model_name
 
     def _load_model(self):
         formatted_args = ""
@@ -399,7 +391,7 @@ def _saved_file_path(self):
         saved_file_name = (
             secure_hash(self.config.predict.name, algorithm="sha384") + ".json"
         )
-        return self._base_path / saved_file_name
+        return self.location / saved_file_name
 
     async def __aenter__(self) -> "VWModel":
         await super().__aenter__()