fiddler-client 3.0.0.dev11__tar.gz → 3.1.0.dev1__tar.gz

{fiddler-client-3.0.0.dev11/fiddler_client.egg-info → fiddler-client-3.1.0.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: fiddler-client
-Version: 3.0.0.dev11
+Version: 3.1.0.dev1
 Summary: Python client for Fiddler Platform
 Home-page: https://fiddler.ai
 Author: Fiddler Labs

fiddler-client-3.1.0.dev1/fiddler/VERSION

@@ -0,0 +1 @@
+3.1.0.dev1

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/entities/events.py

@@ -1,11 +1,13 @@
 from __future__ import annotations
 
+import tempfile
 from pathlib import Path
-from typing import Any, Iterator
+from typing import Any
 from uuid import UUID
 
 import pandas as pd
 from requests import Response
+from tqdm import tqdm
 
 from fiddler.configs import STREAM_EVENT_LIMIT
 from fiddler.connection import ConnectionMixin
@@ -28,21 +30,6 @@ class EventPublisher(ConnectionMixin):
         """
         self.model_id = model_id
 
-    def _get_stream_chunks(
-        self, source: pd.DataFrame | list[dict[str, Any]]
-    ) -> Iterator[list[dict[str, Any]]]:
-        """Chunk the source based on stream limit"""
-
-        for i in range(0, len(source), self.STREAM_LIMIT):
-            chunk = source[i : i + self.STREAM_LIMIT]
-
-            if isinstance(chunk, pd.DataFrame):
-                events = chunk.to_dict('records')
-            else:
-                events = chunk
-
-            yield events
-
     @handle_api_error
     def publish(
         self,
@@ -56,13 +43,13 @@ class EventPublisher(ConnectionMixin):
 
         :param source: one of:
             Path or str path: path for data file.
-            list[dict]: list of event dicts (max 1000) EnvType.PRE_PRODUCTION is not supported
-            dataframe: events dataframe. EnvType.PRE_PRODUCTION not supported.
+            dataframe: events dataframe.
+            list[dict]: list of event dicts EnvType.PRE_PRODUCTION is not supported
         :param environment: Either EnvType.PRE_PRODUCTION or EnvType.PRODUCTION
         :param dataset_name: Name of the dataset. Not supported for EnvType.PRODUCTION
         :param update: flag indicating if the events are updates to previously published rows
 
-        :return: list[UUID] for list of dicts or dataframe source and Job object for file path source.
+        :return: list[UUID] for list of dicts source and Job object for file path or dataframe source.
         """
         if isinstance(source, (str, Path)):
             return self._publish_file(
@@ -72,20 +59,30 @@ class EventPublisher(ConnectionMixin):
                 update=update,
             )
 
-        if isinstance(source, (list, pd.DataFrame)):
+        if isinstance(source, pd.DataFrame):
+            with tempfile.NamedTemporaryFile(suffix='.csv') as temp_file:
+                source.to_csv(temp_file.name, index=False)
+                return self._publish_file(
+                    source=temp_file.name,
+                    environment=environment,
+                    dataset_name=dataset_name,
+                    update=update,
+                )
+
+        elif isinstance(source, list):
             return self._publish_stream(source=source, update=update)
 
         raise ValueError(f'Unsupported source - {type(source)}')
 
     def _publish_stream(
         self,
-        source: list[dict[str, Any]] | pd.DataFrame,
+        source: list[dict[str, Any]],
         update: bool = False,
     ) -> list[UUID]:
         event_ids = []
-        for events in self._get_stream_chunks(source=source):
+        for i in tqdm(range(0, len(source), self.STREAM_LIMIT)):
             response = self._publish_call(
-                source=EventsSource(events=events),
+                source=EventsSource(events=source[i : i + self.STREAM_LIMIT]),
                 environment=EnvType.PRODUCTION,
                 dataset_name=None,
                 update=update,

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/entities/model.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import builtins
+import copy
 import typing
 from dataclasses import dataclass
 from datetime import datetime
@@ -52,6 +53,7 @@ class Model(
         project_id: UUID | str,
         schema: ModelSchema,
         spec: ModelSpec,
+        version: str | None = None,
         input_type: str = ModelInputType.TABULAR,
         task: str = ModelTask.NOT_SET,
         task_params: ModelTaskParams | None = None,
@@ -63,6 +65,7 @@ class Model(
     ) -> None:
         """Construct a model instance"""
         self.name = name
+        self.version = version
         self.project_id = project_id
         self.schema = schema
         self.input_type = input_type
@@ -101,6 +104,7 @@ class Model(
         # Initialize
         instance = cls(
             name=resp_obj.name,
+            version=resp_obj.version,
             schema=resp_obj.schema_,
             spec=resp_obj.spec,
             project_id=resp_obj.project.id,
@@ -141,6 +145,7 @@ class Model(
         fields = [
             'id',
             'name',
+            'version',
             'spec',
             'input_type',
             'task',
@@ -188,12 +193,20 @@ class Model(
 
     @classmethod
     @handle_api_error
-    def from_name(cls, name: str, project_id: UUID | str) -> Model:
+    def from_name(
+        cls,
+        name: str,
+        project_id: UUID | str,
+        version: str | None = None,
+        latest: bool = False,
+    ) -> Model:
         """
         Get the model instance from model name within a project.
 
         :param name: Model name
         :param project_id: Project identifier
+        :param version: Version name
+        :param latest: Whether to fetch the latest version of the model or first version
         :return: Model instance
         """
         _filter = QueryCondition(
@@ -207,8 +220,15 @@ class Model(
             ]
         )
 
+        if version:
+            _filter.add_rule(
+                QueryRule(field='version', operator=OperatorType.EQUAL, value=version)
+            )
+
+        ordering = '-created_at' if latest else 'created_at'
         response = cls._client().get(
-            url=cls._get_url(), params={'filter': _filter.json()}
+            url=cls._get_url(),
+            params={'filter': _filter.json(), 'limit': 1, 'ordering': ordering},
         )
 
         if response.json()['data']['total'] == 0:
@@ -219,22 +239,27 @@ class Model(
     @handle_api_error
     def create(self) -> Model:
         """Create a new model."""
+        payload = {
+            'name': self.name,
+            'project_id': str(self.project_id),
+            'schema': self.schema.dict(),
+            'spec': self.spec.dict(),
+            'input_type': self.input_type,
+            'task': self.task,
+            'task_params': self.task_params.dict(),
+            'description': self.description,
+            'event_id_col': self.event_id_col,
+            'event_ts_col': self.event_ts_col,
+            'event_ts_format': self.event_ts_format,
+            'xai_params': self.xai_params.dict(),
+        }
+
+        if self.version:
+            payload['version'] = self.version
+
         response = self._client().post(
             url=self._get_url(),
-            data={
-                'name': self.name,
-                'project_id': str(self.project_id),
-                'schema': self.schema.dict(),
-                'spec': self.spec.dict(),
-                'input_type': self.input_type,
-                'task': self.task,
-                'task_params': self.task_params.dict(),
-                'description': self.description,
-                'event_id_col': self.event_id_col,
-                'event_ts_col': self.event_ts_col,
-                'event_ts_format': self.event_ts_format,
-                'xai_params': self.xai_params.dict(),
-            },
+            data=payload,
         )
         self._refresh_from_response(response=response)
         return self
@@ -243,6 +268,7 @@ class Model(
     def update(self) -> None:
         """Update an existing model."""
         body: dict[str, Any] = {
+            'version': self.version,
             'xai_params': self.xai_params.dict(),
             'description': self.description,
             'event_id_col': self.event_id_col,
@@ -256,13 +282,13 @@ class Model(
     @classmethod
     @handle_api_error
     def list(
-        cls,
-        project_id: UUID | str,
+        cls, project_id: UUID | str, name: str | None = None
     ) -> Iterator[ModelCompact]:
         """
         Get a list of all models with the given filters
 
         :param project_id: Project identifier
+        :param name: Model name, use this for listing all the versions of a model
         :return: ModelCompact iterator
         """
         _filter = QueryCondition(
@@ -273,11 +299,34 @@ class Model(
             ]
         )
 
+        if name:
+            _filter.add_rule(
+                QueryRule(field='name', operator=OperatorType.EQUAL, value=name)
+            )
+
         params = {'filter': _filter.json()}
 
         for model in cls._paginate(url=cls._get_url(), params=params):
             yield ModelCompact(id=model['id'], name=model['name'])
 
+    def duplicate(self, version: str | None = None) -> Model:
+        """
+        Duplicate the model instance with the given version name.
+
+        This call will not save the model on server. After making changes to the model
+        instance call .create() to add the model version to Fiddler Platform.
+
+        :param version: Version name for the new instance
+        :return: Model instance
+        """
+        self_copy = copy.deepcopy(self)
+        self_copy.id = None
+
+        if version:
+            self_copy.version = version
+
+        return self_copy
+
     @property
     def datasets(self) -> Iterator[Dataset]:
         """Fetch all the datasets of this model"""
@@ -316,6 +365,7 @@ class Model(
         name: str,
         project_id: UUID | str,
         spec: ModelSpec | None = None,
+        version: str | None = None,
         input_type: str = ModelInputType.TABULAR,
         task: str = ModelTask.NOT_SET,
         task_params: ModelTaskParams | None = None,
@@ -334,6 +384,7 @@ class Model(
         :param name: Model name
         :param project_id: Project identifier
         :param spec: ModelSpec instance
+        :param version: Model version name
         :param input_type: Model input task type
         :param task: Model task like Regression, Binary classification etc.
         :param task_params: Parameters based on the task type
@@ -354,6 +405,7 @@ class Model(
 
         return Model(
             name=name,
+            version=version,
             schema=resp_obj.schema_,
             spec=resp_obj.spec,
             project_id=project_id,

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/schemas/model.py

@@ -22,6 +22,7 @@ class ModelCompactResp(BaseModel):
 class ModelResp(BaseModel):
     id: UUID
     name: str
+    version: str
     input_type: str
     task: str
     task_params: ModelTaskParams

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/tests/apis/test_events.py

@@ -100,52 +100,74 @@ def test_publish_file() -> None:
 
 
 @responses.activate
-def test_publish_stream() -> None:
+def test_publish_df() -> None:
     responses.get(
         url=f'{URL}/v3/models/{MODEL_ID}',
         json=MODEL_API_RESPONSE_200,
     )
     model = Model.get(id_=MODEL_ID)
 
+    responses.post(
+        url=f'{URL}/v3/files/upload',
+        json=SINGLE_UPLOAD_200_RESPONSE,
+    )
+    responses.get(
+        url=f'{URL}/v3/jobs/{JOB_ID}',
+        json=PUBLISH_JOB_RESPONSE,
+    )
+
     responses.post(
         url=f'{URL}/v3/events',
-        json=STREAM_PUBLISH_202_API_RESPONSE,
+        json=FILE_PUBLISH_202_API_RESPONSE,
     )
     publish_response = model.publish(
-        source=df,
-        environment=EnvType.PRODUCTION,
+        source=df, environment=EnvType.PRE_PRODUCTION, dataset_name='dataset_1'
     )
 
-    assert isinstance(publish_response, list)
-    assert len(publish_response) == df.shape[0]
+    assert isinstance(publish_response, Job)
+
+    FILE_PUBLISH_202_API_RESPONSE['data']['job']['name'] = 'Publish events'
 
     publish_response = model.publish(
-        source=df.to_dict('records'),
+        source=df,
         environment=EnvType.PRODUCTION,
     )
 
-    assert isinstance(publish_response, list)
-    assert len(publish_response) == df.shape[0]
+    assert isinstance(publish_response, Job)
+
+    FILE_PUBLISH_202_API_RESPONSE['data']['job']['name'] = 'Publish events'
 
 
 @responses.activate
-def test_publish_update() -> None:
+def test_publish_stream() -> None:
     responses.get(
         url=f'{URL}/v3/models/{MODEL_ID}',
         json=MODEL_API_RESPONSE_200,
     )
     model = Model.get(id_=MODEL_ID)
 
-    responses.patch(
+    responses.post(
         url=f'{URL}/v3/events',
         json=STREAM_PUBLISH_202_API_RESPONSE,
     )
+
     publish_response = model.publish(
-        source=df, environment=EnvType.PRODUCTION, update=True
+        source=df.to_dict('records'),
+        environment=EnvType.PRODUCTION,
     )
+
     assert isinstance(publish_response, list)
     assert len(publish_response) == df.shape[0]
 
+
+@responses.activate
+def test_publish_update() -> None:
+    responses.get(
+        url=f'{URL}/v3/models/{MODEL_ID}',
+        json=MODEL_API_RESPONSE_200,
+    )
+    model = Model.get(id_=MODEL_ID)
+
     responses.post(
         url=f'{URL}/v3/files/upload',
         json=SINGLE_UPLOAD_200_RESPONSE,
@@ -167,3 +189,8 @@ def test_publish_update() -> None:
     )
 
     assert isinstance(publish_response, Job)
+
+    publish_response = model.publish(
+        source=df, environment=EnvType.PRODUCTION, update=True
+    )
+    assert isinstance(publish_response, Job)

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/tests/apis/test_generate_model.py

@@ -182,7 +182,12 @@ def test_generate_model_with_parquet_file() -> None:
 
         DF.to_csv(temp_file.name, index=False)
 
+        version = 'v1.0'
         model = Model.from_data(
-            source=temp_file.name, name='model1', project_id=PROJECT_ID
+            source=temp_file.name,
+            name='model1',
+            version=version,
+            project_id=PROJECT_ID,
         )
         assert isinstance(model, Model)
+        assert model.version == version

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/tests/apis/test_model.py

@@ -18,6 +18,7 @@ from fiddler.tests.constants import (
     MODEL_DEPLOYMENT_ID,
     MODEL_ID,
     MODEL_NAME,
+    MODEL_VERSION,
     ORG_ID,
     ORG_NAME,
     PROJECT_ID,
@@ -153,6 +154,7 @@ API_RESPONSE_200 = {
     'data': {
         'id': MODEL_ID,
         'name': MODEL_NAME,
+        'version': MODEL_VERSION,
         'project': {
             'id': PROJECT_ID,
             'name': PROJECT_NAME,
@@ -308,6 +310,7 @@ def test_add_model_success() -> None:
     assert isinstance(model, Model)
     assert model.id == UUID(MODEL_ID)
     assert model.name == MODEL_NAME
+    assert model.version == MODEL_VERSION
 
 
 @responses.activate
@@ -350,7 +353,9 @@ def test_get_model_not_found() -> None:
 @responses.activate
 def test_model_from_name_success() -> None:
     params = {
-        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}'
+        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}',
+        'limit': 1,
+        'ordering': 'created_at',
     }
     responses.get(
         url=f'{URL}/v3/models',
@@ -365,6 +370,48 @@ def test_model_from_name_success() -> None:
     assert isinstance(model, Model)
 
 
+@responses.activate
+def test_model_from_name_with_version_success() -> None:
+    params = {
+        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}, {"field": "version", "operator": "equal", "value": "v1"}]}',
+        'limit': 1,
+        'ordering': 'created_at',
+    }
+    responses.get(
+        url=f'{URL}/v3/models',
+        json=API_RESPONSE_FROM_NAME,
+        match=[matchers.query_param_matcher(params)],
+    )
+    responses.get(
+        url=f'{URL}/v3/models/{MODEL_ID}',
+        json=API_RESPONSE_200,
+    )
+    model = Model.from_name(
+        name=MODEL_NAME, version=MODEL_VERSION, project_id=PROJECT_ID
+    )
+    assert isinstance(model, Model)
+
+
+@responses.activate
+def test_model_from_name_latest_success() -> None:
+    params = {
+        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}',
+        'limit': 1,
+        'ordering': '-created_at',
+    }
+    responses.get(
+        url=f'{URL}/v3/models',
+        json=API_RESPONSE_FROM_NAME,
+        match=[matchers.query_param_matcher(params)],
+    )
+    responses.get(
+        url=f'{URL}/v3/models/{MODEL_ID}',
+        json=API_RESPONSE_200,
+    )
+    model = Model.from_name(name=MODEL_NAME, project_id=PROJECT_ID, latest=True)
+    assert isinstance(model, Model)
+
+
 @responses.activate
 def test_model_from_name_not_found() -> None:
     resp = API_RESPONSE_FROM_NAME.copy()
@@ -373,7 +420,9 @@ def test_model_from_name_not_found() -> None:
     resp['data']['items'] = []
 
     params = {
-        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}'
+        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}',
+        'limit': 1,
+        'ordering': 'created_at',
     }
     responses.get(
         url=f'{URL}/v3/models',
@@ -385,9 +434,35 @@ def test_model_from_name_not_found() -> None:
         Model.from_name(name=MODEL_NAME, project_id=PROJECT_ID)
 
 
+@responses.activate
+def test_model_from_name_with_version_not_found() -> None:
+    resp = API_RESPONSE_FROM_NAME.copy()
+    resp['data']['total'] = 0
+    resp['data']['item_count'] = 0
+    resp['data']['items'] = []
+
+    params = {
+        'filter': '{"condition": "AND", "rules": [{"field": "name", "operator": "equal", "value": "bank_churn"}, {"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}, {"field": "version", "operator": "equal", "value": "v1"}]}',
+        'limit': 1,
+        'ordering': 'created_at',
+    }
+    responses.get(
+        url=f'{URL}/v3/models',
+        json=resp,
+        match=[matchers.query_param_matcher(params)],
+    )
+
+    with pytest.raises(NotFound):
+        Model.from_name(name=MODEL_NAME, version=MODEL_VERSION, project_id=PROJECT_ID)
+
+
 @responses.activate
 def test_model_list_success() -> None:
-    params = {'project_id': str(PROJECT_ID), 'limit': 50, 'offset': 0}
+    params = {
+        'filter': '{"condition": "AND", "rules": [{"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}',
+        'limit': 50,
+        'offset': 0,
+    }
     responses.get(
         url=f'{URL}/v3/models',
         json=LIST_API_RESPONSE,
@@ -399,9 +474,9 @@ def test_model_list_success() -> None:
 
 
 @responses.activate
-def test_model_list_success() -> None:
+def test_model_list_with_name() -> None:
     params = {
-        'filter': '{"condition": "AND", "rules": [{"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}]}',
+        'filter': '{"condition": "AND", "rules": [{"field": "project_id", "operator": "equal", "value": "1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1"}, {"field": "name", "operator": "equal", "value": "bank_churn"}]}',
         'limit': 50,
         'offset': 0,
     }
@@ -410,7 +485,7 @@ def test_model_list_success() -> None:
         json=LIST_API_RESPONSE,
         match=[matchers.query_param_matcher(params)],
     )
-    models = Model.list(project_id=PROJECT_ID)
+    models = Model.list(project_id=PROJECT_ID, name=MODEL_NAME)
     for model in models:
         assert isinstance(model, ModelCompact)
 
@@ -471,6 +546,18 @@ def test_update_model_success(mocker) -> None:
     model.update()
     assert model.description == 'Test model update'
 
+    # update version
+    new_version = 'v2'
+    mocker.patch.dict(API_RESPONSE_200['data'], {'version': new_version})
+
+    responses.patch(
+        url=f'{URL}/v3/models/{MODEL_ID}',
+        json=API_RESPONSE_200,
+    )
+    model.version = new_version
+    model.update()
+    assert model.version == new_version
+
 
 @responses.activate
 def test_update_model_not_found() -> None:
@@ -556,3 +643,25 @@ def test_model_schema_dunder_methods() -> None:
     # Delete method - KeyError
     with pytest.raises(KeyError):
         del schema['foo']
+
+
+@responses.activate
+def test_model_duplicate() -> None:
+    responses.get(
+        url=f'{URL}/v3/models/{MODEL_ID}',
+        json=API_RESPONSE_200,
+    )
+    model = Model.get(id_=MODEL_ID)
+
+    assert model.id == UUID(MODEL_ID)
+
+    # Without version parameter
+    model_copy = model.duplicate()
+    assert model_copy.id is None
+    assert model_copy.version == model.version
+
+    # With version parameter
+    new_version = 'v2'
+    model_copy = model.duplicate(version=new_version)
+    assert model_copy.id is None
+    assert model_copy.version == new_version

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/tests/apis/test_xai.py

@@ -16,6 +16,7 @@ from fiddler.tests.constants import (
     JOB_ID,
     MODEL_ID,
     MODEL_NAME,
+    MODEL_VERSION,
     ORG_ID,
     ORG_NAME,
     OUTPUT_DIR,
@@ -70,6 +71,7 @@ MODEL_API_RESPONSE_200 = {
     'data': {
         'id': MODEL_ID,
         'name': MODEL_NAME,
+        'version': MODEL_VERSION,
         'project': {
             'id': PROJECT_ID,
             'name': PROJECT_NAME,

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1}/fiddler/tests/constants.py

@@ -10,6 +10,7 @@ PROJECT_NAME = 'bank_churn'
 PROJECT_ID = '1531bfd9-2ca2-4a7b-bb5a-136c8da09ca1'
 MODEL_NAME = 'bank_churn'
 MODEL_ID = '4531bfd9-2ca2-4a7b-bb5a-136c8da09ca2'
+MODEL_VERSION = 'v1'
 USER_ID = '5531bfd9-2ca2-4a7b-bb5a-136c8da09ca3'
 USER_NAME = 'foo'
 USER_EMAIL = 'foo@fiddler.ai'

{fiddler-client-3.0.0.dev11 → fiddler-client-3.1.0.dev1/fiddler_client.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: fiddler-client
-Version: 3.0.0.dev11
+Version: 3.1.0.dev1
 Summary: Python client for Fiddler Platform
 Home-page: https://fiddler.ai
 Author: Fiddler Labs

fiddler-client-3.0.0.dev11/fiddler/VERSION

@@ -1 +0,0 @@
-3.0.0-dev11