wizata-dsapi 1.3.19__tar.gz → 1.3.21__tar.gz

{wizata_dsapi-1.3.19/wizata_dsapi.egg-info → wizata_dsapi-1.3.21}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: wizata_dsapi
-Version: 1.3.19
+Version: 1.3.21
 Summary: Wizata Data Science Toolkit
 Author: Wizata S.A.
 Author-email: info@wizata.com

{wizata_dsapi-1.3.19 → wizata_dsapi-1.3.21}/wizata_dsapi/mlmodel.py

@@ -1,10 +1,5 @@
-import json
-import uuid
-
 from typing import List, Iterator
-
 from .api_dto import ApiDto
-from enum import Enum
 
 
 def get_bool(obj, name: str):
@@ -14,6 +9,118 @@ def get_bool(obj, name: str):
         return bool(obj[name])
 
 
+class ModelInfo:
+    """
+    define a pointer to a machine learning model.
+
+    :ivar model_type str: mlflow model flavour
+    """
+
+    def __init__(self,
+                 key: str,
+                 twin_hardware_id: str = None,
+                 property_name: str = None,
+                 alias: str = None,
+                 model_type: str = None,
+                 file_format: str = 'pkl',
+                 source: str = 'wizata'
+                 ):
+        self.key = key
+        self.twin_hardware_id = twin_hardware_id
+        self.property_name = property_name
+        self.alias = alias
+        self.file_format = file_format
+        self.source = source
+        self.model_type = model_type
+        self.trained_model = None
+        self.scaler = None
+
+    def identifier(self, include_alias: bool = False) -> str:
+        """
+        returns the complete string identifier for this model.
+        :param include_alias: include the alias pointer or leave it to target the default version.
+        :return: complete identifier of a model.
+        """
+        if self.key is None:
+            raise KeyError('please specific a model key')
+        identifier = self.key
+
+        if self.twin_hardware_id is not None:
+            identifier += f".{self.twin_hardware_id}"
+
+        if self.property_name is not None:
+            identifier += f".{self.property_name}"
+
+        if include_alias and self.alias is not None:
+            identifier += f"@{self.alias}"
+
+        return identifier
+
+    def to_json(self):
+        """
+        convert this entity in a dict that can be json serializable
+        :return: dict
+        """
+        obj = {
+            "key": self.key,
+            "file_format": self.file_format,
+            "source": self.source
+        }
+        if self.twin_hardware_id is not None:
+            obj["twin_hardware_id"] = str(self.twin_hardware_id)
+        if self.property_name is not None:
+            obj["property_name"] = self.property_name
+        if self.alias is not None:
+            obj["alias"] = self.alias
+        if self.model_type is not None:
+            obj["model_type"] = self.model_type
+        if self.file_format is not None:
+            obj["file_format"] = self.file_format
+        if self.source is not None:
+            obj["source"] = self.source
+        return obj
+
+    def from_json(self, obj):
+        """
+        load this entity from a dict
+        """
+        if "key" in obj.keys():
+            self.key = obj["key"]
+        if "twin_hardware_id" in obj.keys():
+            self.twin_hardware_id = obj["twin_hardware_id"]
+        if "property_name" in obj.keys():
+            self.property_name = obj["property_name"]
+        if "alias" in obj.keys():
+            self.alias = obj["alias"]
+        if "model_type" in obj.keys():
+            self.model_type = obj["model_type"]
+        if "file_format" in obj.keys():
+            self.key = obj["file_format"]
+        if "source" in obj.keys():
+            self.source = obj["source"]
+
+
+class ModelList:
+    """
+    used to conveniently manipulate a list of models.
+    """
+
+    def __init__(self):
+        self.models: List[ModelInfo] = []
+
+    def __iter__(self) -> Iterator[ModelInfo]:
+        return iter(self.models)
+
+    def __len__(self) -> int:
+        return len(self.models)
+
+    def __getitem__(self, index: int) -> ModelInfo:
+        return self.models[index]
+
+    def append(self, model: ModelInfo):
+        self.models.append(model)
+
+
 class MLModelConfig(ApiDto):
     """
     a model config defines execution properties within a pipeline.
@@ -87,6 +194,25 @@ class MLModelConfig(ApiDto):
         self.model_type = model_type
         self.model_alias = model_alias
 
+    def create_model_info(self, hardware_id: str = None) -> ModelInfo:
+        """
+        create model info corresponding to the configuration.
+        :param hardware_id: provide a hardware id for this model if by_twin.
+        :return:
+        """
+        if self.by_twin and hardware_id is None:
+            raise ValueError('hardware_id is required if by_twin to create a model info')
+        model_info = ModelInfo(
+            key=self.model_key,
+            twin_hardware_id=hardware_id,
+            property_name=self.property_name,
+            source=self.source,
+            alias=self.model_alias,
+            file_format=self.model_format,
+            model_type=self.model_type
+        )
+        return model_info
+
     def from_json(self, obj):
 
         # Managed deprecated fields
@@ -230,228 +356,3 @@ class MLModelConfig(ApiDto):
         else:
             raise TypeError(f'unsupported target_feat type {self.target_feat.__class__.__name__}')
 
-
-class ModelInfo:
-    """
-    define a pointer to a machine learning model.
-    """
-
-    def __init__(self,
-                 key: str,
-                 twin_hardware_id: str = None,
-                 property_name: str = None,
-                 alias: str = None,
-                 file_format: str = 'pkl',
-                 source: str = 'wizata'
-                 ):
-        self.key = key
-        self.twin_hardware_id = twin_hardware_id
-        self.property_name = property_name
-        self.alias = alias
-        self.file_format = file_format
-        self.source = source
-        self.trained_model = None
-        self.scaler = None
-
-    def identifier(self, include_alias: bool = False) -> str:
-        """
-        returns the complete string identifier for this model.
-        :param include_alias: include the alias pointer or leave it to target the default version.
-        :return: complete identifier of a model.
-        """
-        if self.key is None:
-            raise KeyError('please specific a model key')
-        identifier = self.key
-
-        if self.twin_hardware_id is not None:
-            identifier += f".{self.twin_hardware_id}"
-
-        if self.property_name is not None:
-            identifier += f".{self.property_name}"
-
-        if include_alias and self.alias is not None:
-                identifier += f"@{self.alias}"
-
-        return identifier
-
-    def to_json(self):
-        """
-        convert this entity in a dict that can be json serializable
-        :return: dict
-        """
-        obj = {
-            "key": self.key,
-            "file_format": self.file_format,
-            "source": self.source
-       }
-        if self.twin_hardware_id is not None:
-            obj["twin_hardware_id"] = str(self.twin_hardware_id)
-        if self.property_name is not None:
-            obj["property_name"] = self.property_name
-        if self.alias is not None:
-            obj["alias"] = self.alias
-        if self.file_format is not None:
-            obj["file_format"] = self.file_format
-        if self.source is not None:
-            obj["source"] = self.source
-        return obj
-
-    def from_json(self, obj):
-        """
-        load this entity from a dict
-        """
-        if "key" in obj.keys():
-            self.key = obj["key"]
-        if "twin_hardware_id" in obj.keys():
-            self.twin_hardware_id = obj["twin_hardware_id"]
-        if "property_name" in obj.keys():
-            self.property_name = obj["property_name"]
-        if "alias" in obj.keys():
-            self.alias = obj["alias"]
-        if "file_format" in obj.keys():
-            self.key = obj["file_format"]
-        if "source" in obj.keys():
-            self.source = obj["source"]
-
-
-class ModelList:
-    """
-    used to conveniently manipulate a list of models.
-    """
-
-    def __init__(self):
-        self.models: List[ModelInfo] = []
-
-    def __iter__(self) -> Iterator[ModelInfo]:
-        return iter(self.models)
-
-    def __len__(self) -> int:
-        return len(self.models)
-
-    def __getitem__(self, index: int) -> ModelInfo:
-        return self.models[index]
-
-    def append(self, model: ModelInfo):
-        self.models.append(model)
-
-
-class MLModel(ApiDto):
-    """
-    used to define a stored machine learning model within Wizata.
-    - deprecated -
-    """
-
-    @classmethod
-    def route(cls):
-        return "mlmodels"
-
-    @classmethod
-    def from_dict(cls, data):
-        obj = MLModel()
-        obj.from_json(data)
-        return obj
-
-    @classmethod
-    def get_type(cls):
-        return "pickle"
-
-    def __init__(self,
-                 model_id: str = None,
-                 generated_by_id=None,
-                 exact_names=True,
-                 exact_numbers=True,
-                 key=None):
-        self.model_id = model_id
-        self.key = key
-
-        self.generatedById = generated_by_id
-
-        self.status = 'draft'
-        self.input_columns = []
-        self.output_columns = []
-
-        self.needExactColumnNumbers = exact_numbers
-        self.needExactColumnNames = exact_names
-        self.has_anomalies = False
-        self.label_counts = 0
-        self.has_target_feat = False
-
-        self.trained_model = None
-        self.scaler = None
-
-        self.experimentId = None
-
-    def api_id(self) -> str:
-        return str(self.model_id).upper()
-
-    def endpoint(self) -> str:
-        return "MLModels"
-
-    def to_json(self, target: str = None):
-        obj = {"id": str(self.model_id),
-               "status": str(self.status),
-               "needExactColumnNames": str(self.needExactColumnNames),
-               "needExactColumnNumbers": str(self.needExactColumnNumbers),
-               "hasAnomalies": str(self.has_anomalies),
-               "hasTargetFeat": str(self.has_target_feat),
-               "labelCount": str(self.label_counts)
-               }
-        if self.key is not None:
-            obj["key"] = str(self.key)
-        if self.generatedById is not None:
-            obj["generatedById"] = self.generatedById
-        if self.input_columns is not None:
-            obj["inputColumns"] = json.dumps(list(self.input_columns))
-        if self.output_columns is not None:
-            obj["outputColumns"] = json.dumps(list(self.output_columns))
-        if self.experimentId is not None:
-            obj["experimentId"] = str(self.experimentId)
-        return obj
-
-    def from_json(self, obj):
-        if "id" in obj.keys():
-            self.model_id = obj["id"]
-        if "key" in obj.keys() and obj["key"] is not None:
-            self.key = obj["key"]
-        if "generatedById" in obj.keys() and obj["generatedById"] is not None:
-            self.generatedById = int(obj["generatedById"])
-        if "experimentId" in obj.keys() and obj["experimentId"] is not None:
-            self.experimentId = uuid.UUID(obj["experimentId"])
-        if "status" in obj.keys():
-            self.status = str(obj["status"]).lower()
-        if "inputColumns" in obj.keys():
-            self.input_columns = json.loads(obj["inputColumns"])
-        if "outputColumns" in obj.keys():
-            self.output_columns = json.loads(obj["outputColumns"])
-        if "labelCount" in obj.keys():
-            self.label_counts = int(obj["labelCount"])
-        if "hasAnomalies" in obj.keys():
-            if isinstance(obj["hasAnomalies"], str) and obj["hasAnomalies"].lower() == "false":
-                self.has_anomalies = False
-            else:
-                self.has_anomalies = bool(obj["hasAnomalies"])
-        if "hasTargetFeat" in obj.keys():
-            if isinstance(obj["hasTargetFeat"], str) and obj["hasTargetFeat"].lower() == "false":
-                self.has_target_feat = False
-            else:
-                self.has_target_feat = bool(obj["hasTargetFeat"])
-        if "needExactColumnNumbers" in obj.keys():
-            if isinstance(obj["needExactColumnNumbers"], str) and obj["needExactColumnNumbers"].lower() == "false":
-                self.needExactColumnNumbers = False
-            else:
-                self.needExactColumnNumbers = bool(obj["needExactColumnNumbers"])
-        if "needExactColumnNames" in obj.keys():
-            if isinstance(obj["needExactColumnNames"], str) and obj["needExactColumnNames"].lower() == "false":
-                self.needExactColumnNames = False
-            else:
-                self.needExactColumnNames = bool(obj["needExactColumnNames"])
-
-    def get_sample_payload(self):
-        pl_columns = {"timestamp": "[timestamp]"}
-        for hardwareId in self.input_columns:
-            pl_columns[hardwareId] = "[" + hardwareId + "]"
-        pl_json = {
-            "id": str(self.model_id),
-            "dataset": pl_columns
-        }
-        return pl_json

wizata_dsapi-1.3.21/wizata_dsapi/version.py

@@ -0,0 +1 @@
+__version__ = "1.3.21"

{wizata_dsapi-1.3.19 → wizata_dsapi-1.3.21}/wizata_dsapi/wizata_dsapi_client.py

@@ -23,7 +23,7 @@ from .api_dto import ApiDto, VarType
 from .business_label import BusinessLabel
 from .plot import Plot
 from .request import Request
-from .mlmodel import MLModel
+from .mlmodel import ModelInfo, ModelList
 from .experiment import Experiment
 from .script import Script
 from .execution import Execution, ExecutionStatus
@@ -590,15 +590,6 @@ class WizataDSAPIClient(ApiInterface):
                 return obj.script_id
             else:
                 raise self.__raise_error(response)
-        if isinstance(obj, MLModel):
-            response = requests.put(self.__url() + "mlmodels/",
-                                    headers=self.__header(),
-                                    data=pickle.dumps(obj))
-            if response.status_code == 200:
-                obj.model_id = uuid.UUID(response.json()['id'])
-                return obj.model_id
-            else:
-                raise self.__raise_error(response)
         if isinstance(obj, Template):
             return self.upsert_template(obj.key, obj.name)
         if isinstance(obj, Pipeline):
@@ -847,13 +838,6 @@ class WizataDSAPIClient(ApiInterface):
                 execution.script = self.get(script_name=script)
             elif isinstance(script, Script):
                 execution.script = script
-        if ml_model is not None:
-            if isinstance(ml_model, uuid.UUID) or (isinstance(ml_model, str) and is_valid_uuid(ml_model)):
-                execution.ml_model = MLModel(ml_model)
-            elif isinstance(ml_model, str):
-                execution.ml_model = self.get(model_key=ml_model)
-            elif isinstance(ml_model, MLModel):
-                execution.ml_model = ml_model
         if image is not None:
             execution.pipeline_image_id = image
         if isAnomalyDetection:
@@ -899,9 +883,6 @@ class WizataDSAPIClient(ApiInterface):
                 if "plots" in obj.keys():
                     for plot in obj["plots"]:
                         result_execution.plots.append(self.get(Plot(plot_id=plot["id"])))
-                if "models" in obj.keys():
-                    for mlmodel in obj["models"]:
-                        result_execution.models.append(self.get(MLModel(model_id=mlmodel["id"])))
                 if "resultDataframe" in obj.keys() and obj["resultDataframe"]["id"] is not None:
                     result_execution.output_ds_dataframe = self.get(DSDataFrame(df_id=obj["resultDataframe"]["id"]))
 
@@ -1104,7 +1085,7 @@ class WizataDSAPIClient(ApiInterface):
                        scaler=None,
                        has_anomalies: bool = False,
                        has_target_feat: bool = False,
-                       experiment_key = None) -> tuple[MLModel, pandas.DataFrame]:
+                       experiment_key = None) -> tuple[ModelInfo, pandas.DataFrame]:
         """
         Register a Machine Learning model to Wizata.
         Model is tested by the API against a sample dataframe.
@@ -1124,15 +1105,13 @@ class WizataDSAPIClient(ApiInterface):
         elif df is None:
             raise ValueError('A sample dataframe must be provided')
 
-        # Create a ML Model object
-        ml_model = MLModel()
+        ml_model = ModelInfo(key=model_key)
         ml_model.trained_model = train_model
         if scaler is not None:
             ml_model.scaler = scaler
         ml_model.has_anomalies = has_anomalies
         ml_model.has_target_feat = has_target_feat
         ml_model.input_columns = df.columns
-        ml_model.key = model_key
 
         if experiment_key is not None:
             ml_model.experimentId = self.get(experiment_key=experiment_key).experiment_id
@@ -1141,7 +1120,8 @@ class WizataDSAPIClient(ApiInterface):
             result_df = predict(df, ml_model)
             if result_df is not None:
                 ml_model.status = "valid"
-                self.upsert(ml_model)
+                #TODO: proper method to upload and download from registry to modify this one
+                ## self.upsert(ml_model)
                 return ml_model, result_df
             else:
                 raise RuntimeError('no dataframe was generated by your model while testing predict capabilities')

{wizata_dsapi-1.3.19 → wizata_dsapi-1.3.21/wizata_dsapi.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: wizata_dsapi
-Version: 1.3.19
+Version: 1.3.21
 Summary: Wizata Data Science Toolkit
 Author: Wizata S.A.
 Author-email: info@wizata.com

wizata_dsapi-1.3.19/wizata_dsapi/version.py

@@ -1 +0,0 @@
-__version__ = "1.3.19"