model: Rename directory property to location

CHANGELOG.md

@@ -23,6 +23,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Calls to hashlib now go through helper functions
 - Build docs using `dffml service dev docs`
 - `cached_download/unpack_archive()` are now functions
+- Directory property to Location
+
 ### Fixed
 - Record object key properties are now always strings
 

dffml/base.py

@@ -3,7 +3,6 @@
 they follow a similar API for instantiation and usage.
 """
 import abc
-import ast
 import copy
 import inspect
 import argparse

dffml/high_level.py

@@ -381,7 +381,7 @@ async def train(model, *args: Union[BaseSource, Record, Dict[str, Any]]):
     ...         Feature("Years", int, 1),
     ...     ),
     ...     predict=Feature("Salary", int, 1),
-    ...     directory="tempdir",
+    ...     location="tempdir",
     ... )
     >>>
     >>> async def main():
@@ -443,7 +443,7 @@ async def accuracy(
     ...         Feature("Years", int, 1),
     ...     ),
     ...     predict=Feature("Salary", int, 1),
-    ...     directory="tempdir",
+    ...     location="tempdir",
     ... )
     >>>
     >>> async def main():
@@ -513,7 +513,7 @@ async def predict(
     ...         Feature("Years", int, 1),
     ...     ),
     ...     predict=Feature("Salary", int, 1),
-    ...     directory="tempdir",
+    ...     location="tempdir",
     ... )
     >>>
     >>> async def main():

dffml/model/model.py

@@ -2,7 +2,7 @@
 # Copyright (c) 2019 Intel Corporation
 """
 Model subclasses are responsible for training themselves on records, making
-predictions about the value of a feature in the record, and assessing thei
+predictions about the value of a feature in the record, and assessing their
 prediction accuracy.
 """
 import abc
@@ -29,13 +29,13 @@ class ModelNotTrained(Exception):
 
 @config
 class ModelConfig:
-    directory: str
+    location: str
     features: Features
 
 
 class ModelContext(abc.ABC, BaseDataFlowFacilitatorObjectContext):
     """
-    Abstract base class which should be derived from and implmented using
+    Abstract base class which should be derived from and implemented using
     various machine learning frameworks or concepts.
     """
 
@@ -80,28 +80,28 @@ def __init__(self, config):
         # TODO Just in case its a string. We should make it so that on
         # instantiation of an @config we convert properties to their correct
         # types.
-        directory = getattr(self.config, "directory", None)
-        if isinstance(directory, str):
-            directory = pathlib.Path(directory)
-        if isinstance(directory, pathlib.Path):
-            # to treat "~" as the the home directory rather than a literal
-            directory = directory.expanduser().resolve()
-            self.config.directory = directory
+        location = getattr(self.config, "location", None)
+        if isinstance(location, str):
+            location = pathlib.Path(location)
+        if isinstance(location, pathlib.Path):
+            # to treat "~" as the the home location rather than a literal
+            location = location.expanduser().resolve()
+            self.config.location = location
 
     def __call__(self) -> ModelContext:
-        self._make_config_directory()
+        self._make_config_location()
         return self.CONTEXT(self)
 
-    def _make_config_directory(self):
+    def _make_config_location(self):
         """
-        If the config object for this model contains the directory property
+        If the config object for this model contains the location property
         then create it if it does not exist.
         """
-        directory = getattr(self.config, "directory", None)
-        if directory is not None:
-            directory = pathlib.Path(directory)
-            if not directory.is_dir():
-                directory.mkdir(mode=MODE_BITS_SECURE, parents=True)
+        location = getattr(self.config, "location", None)
+        if location is not None:
+            location = pathlib.Path(location)
+            if not location.is_dir():
+                location.mkdir(mode=MODE_BITS_SECURE, parents=True)
 
 
 class SimpleModelNoContext:
@@ -131,7 +131,7 @@ async def __aenter__(self) -> Model:
         # If we've already entered the model's context once, don't reload
         if self._in_context > 1:
             return self
-        self._make_config_directory()
+        self._make_config_location()
         self.open()
         return self
 
@@ -171,20 +171,20 @@ def close(self):
     def disk_path(self, extention: Optional[str] = None):
         """
         We do this for convenience of the user so they can usually just use the
-        default directory and if they train models with different parameters
+        default location and if they train models with different parameters
         this method transparently to the user creates a filename unique the that
         configuration of the model where data is saved and loaded.
         """
         # Export the config to a dictionary
         exported = self.config._asdict()
-        # Remove the directory from the exported dict
-        if "directory" in exported:
-            del exported["directory"]
+        # Remove the location from the exported dict
+        if "location" in exported:
+            del exported["location"]
         # Replace features with the sorted list of features
         if "features" in exported:
             exported["features"] = dict(sorted(exported["features"].items()))
         # Hash the exported config
-        return pathlib.Path(self.config.directory, "Model",)
+        return pathlib.Path(self.config.location, "Model",)
 
     def applicable_features(self, features):
         usable = []

dffml/model/slr.py

@@ -50,7 +50,7 @@ def best_fit_line(x, y):
 class SLRModelConfig:
     predict: Feature = field("Label or the value to be predicted")
     features: Features = field("Features to train on. For SLR only 1 allowed")
-    directory: pathlib.Path = field("Directory where state should be saved")
+    location: pathlib.Path = field("Location where state should be saved")
 
 
 @entrypoint("slr")
@@ -82,7 +82,7 @@ class SLRModel(SimpleModel):
             -model slr \
             -model-features f1:float:1 \
             -model-predict ans:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename dataset.csv
 
@@ -95,7 +95,7 @@ class SLRModel(SimpleModel):
             -model slr \
             -model-features f1:float:1 \
             -model-predict ans:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename dataset.csv
         1.0
@@ -118,7 +118,7 @@ class SLRModel(SimpleModel):
             -model slr \
             -model-features f1:float:1 \
             -model-predict ans:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename predict.csv
         [

dffml/operation/model.py

@@ -61,7 +61,7 @@ async def model_predict(self, features: Dict[str, Any]) -> Dict[str, Any]:
     >>> slr_model = SLRModel(
     ...     features=Features(Feature("Years", int, 1)),
     ...     predict=Feature("Salary", int, 1),
-    ...     directory="tempdir",
+    ...     location="tempdir",
     ... )
     >>> dataflow = DataFlow(
     ...     operations={

dffml/skel/model/README.md

@@ -18,7 +18,7 @@ sed -i 's/.*setosa,versicolor,virginica/SepalLength,SepalWidth,PetalLength,Petal
 head iris_training.csv
 dffml train \
   -model model_name \
-  -model-directory tempdir \
+  -model-location tempdir \
   -sources csv=iris_training.csv \
   -classifications 0 1 2 \
   -model-features \
@@ -31,7 +31,7 @@ dffml train \
   -log debug
 dffml accuracy \
   -model model_name \
-  -model-directory tempdir \
+  -model-location tempdir \
   -sources csv=iris_training.csv \
   -classifications 0 1 2 \
   -model-features \
@@ -42,7 +42,7 @@ dffml accuracy \
   -log critical
 dffml predict all \
   -model model_name \
-  -model-directory tempdir \
+  -model-location tempdir \
   -sources csv=iris_test.csv \
   -classifications 0 1 2 \
   -model-features \

dffml/skel/model/REPLACE_IMPORT_PACKAGE_NAME/myslr.py

@@ -60,7 +60,7 @@ class MySLRModelConfig:
         "Features to train on (myslr only supports one)"
     )
     predict: Feature = field("Label or the value to be predicted")
-    directory: pathlib.Path = field("Directory where state should be saved")
+    location: pathlib.Path = field("Location where state should be saved")
 
 
 @entrypoint("myslr")
@@ -105,7 +105,7 @@ class MySLRModel(SimpleModel):
             -model myslr \
             -model-features x:float:1 \
             -model-predict y:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename train.csv
 
@@ -118,7 +118,7 @@ class MySLRModel(SimpleModel):
             -model myslr \
             -model-features x:float:1 \
             -model-predict y:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename test.csv
         1.0
@@ -141,7 +141,7 @@ class MySLRModel(SimpleModel):
             -model myslr \
             -model-features x:float:1 \
             -model-predict y:int:1 \
-            -model-directory tempdir \
+            -model-location tempdir \
             -sources f=csv \
             -source-filename predict.csv
         [

dffml/skel/model/examples/example_myslr.py

@@ -6,7 +6,7 @@
 model = MySLRModel(
     features=Features(Feature("x", float, 1)),
     predict=Feature("y", int, 1),
-    directory="tempdir",
+    location="tempdir",
 )
 
 # Train the model

dffml/skel/model/tests/test_model.py

@@ -43,7 +43,7 @@ def setUpClass(cls):
         cls.model = MySLRModel(
             features=Features(Feature("X", float, 1)),
             predict=Feature("Y", float, 1),
-            directory=cls.model_dir.name,
+            location=cls.model_dir.name,
         )
 
     @classmethod

docs/examples/icecream_sales.rst

@@ -142,7 +142,7 @@ and the output being sales.
         -model-hidden 9 18 9 \
         -model-features population:int:1 temperature:float:1 \
         -model-predict sales:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources f=csv \
         -source-filename preprocessed.csv \
         -log debug
@@ -163,7 +163,7 @@ using the accuracy method.
         -model-hidden 9 18 9 \
         -model-features population:int:1 temperature:float:1 \
         -model-predict sales:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources f=csv \
         -source-filename preprocessed.csv \
         -log debug
@@ -188,7 +188,7 @@ for the prediction of sales.
         -model-hidden 9 18 9 \
         -model-features population:int:1 temperature:float:1 \
         -model-predict sales:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources preprocess=df \
         -source-preprocess-dataflow preprocess_ops.json \
         -source-preprocess-features city:str:1 state:str:1 month:int:1 \

docs/tutorials/dataflows/nlp.rst

@@ -107,7 +107,7 @@ We can now use this dataflow to preprocess the data and make it ready to be fed
         -model-clstype int \
         -model-predict sentiment:int:1 \
         -model-classifications 0 1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -model-features embedding:float:[1,10,96] \
         -sources text=df \
         -source-text-dataflow nlp_ops_dataflow.json \
@@ -132,7 +132,7 @@ Assess accuracy:
         -model-clstype int \
         -model-predict sentiment:int:1 \
         -model-classifications 0 1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -model-features embedding:float:[1,10,96] \
         -sources text=df \
         -source-text-dataflow nlp_ops_dataflow.json \
@@ -165,7 +165,7 @@ Make prediction on test data:
         -model-clstype int \
         -model-predict sentiment:int:1 \
         -model-classifications 0 1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -model-features embedding:float:[1,10,96] \
         -sources text=df \
         -source-text-dataflow nlp_ops_dataflow.json \
@@ -273,7 +273,7 @@ We can now use this dataflow to preprocess the data and make it ready to be fed
         -model scikitgnb \
         -model-features extract_array_from_matrix.outputs.result:float:1 \
         -model-predict sentiment:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources text=df \
         -source-text-dataflow nlp_ops_dataflow.json \
         -source-text-features sentence:str:1 \
@@ -290,7 +290,7 @@ Assess accuracy:
         -model scikitgnb \
         -model-features extract_array_from_matrix.outputs.result:float:1 \
         -model-predict sentiment:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources text=df \
         -source-text-dataflow nlp_ops_dataflow.json \
         -source-text-features sentence:str:1 \
@@ -348,7 +348,7 @@ Now we can make prediction on test data:
         -model scikitgnb \
         -model-features extract_array_from_matrix.outputs.result:float:1 \
         -model-predict sentiment:int:1 \
-        -model-directory tempdir \
+        -model-location tempdir \
         -sources temp=json \
         -source-temp-filename test_data_preprocessed.json \
         -pretty

docs/tutorials/models/load.rst

@@ -32,4 +32,4 @@ dataclass using the :py:func:`dataclasses.fields` function.
     <class 'dffml.model.slr.SLRModel'>
         predict: Label or the value to be predicted
         features: Features to train on. For SLR only 1 allowed
-        directory: Directory where state should be saved
+        location: Location where state should be saved

docs/tutorials/models/package.rst

@@ -257,7 +257,7 @@ reference it by it's short name.
         -model myslr \
         -model-features Years:int:1 \
         -model-predict Salary:float:1 \
-        -model-directory modeldir \
+        -model-location modeldir \
         -sources f=csv \
         -source-filename train.csv
 

docs/tutorials/models/slr.rst

@@ -245,7 +245,7 @@ We do the same steps we did with Python, only using the command line interface.
         -model myslr:MySLRModel \
         -model-features Years:int:1 \
         -model-predict Salary:float:1 \
-        -model-directory modeldir \
+        -model-location modeldir \
         -sources f=csv \
         -source-filename train.csv
 
@@ -260,7 +260,7 @@ Now let's make predictions
         -model myslr:MySLRModel \
         -model-features Years:int:1 \
         -model-predict Salary:float:1 \
-        -model-directory modeldir \
+        -model-location modeldir \
         -sources f=csv \
         -source-filename predict.csv
     [
@@ -311,7 +311,7 @@ via the HTTP :ref:`plugin_service_http_api_model` API.
         -models mymodel=myslr:MySLRModel \
         -model-features Years:int:1 \
         -model-predict Salary:float:1 \
-        -model-directory modeldir
+        -model-location modeldir
 
 We can then ask the HTTP service to make predictions, or do training or accuracy
 assessment.