[DEV-2054] added support for new extract and extract_historical funct…

…ions (#332)

* added support for new extract and extract_historical functions

* added support for new extract and extract_historical functions

* Updated the documents

* Updated the documents

* fixed auto lint changes

* fixed auto lint changes

* Removed Featureset as input type from client function docstrings and parameter definition

* Removed Featureset as input type from client function docstrings and parameter definition

* Pushed version

* pr changes

* ran black

* fixed request error

* Added warning about deprecation

docs/examples/api-reference/client.py

@@ -1,9 +1,9 @@
 import unittest
 from datetime import datetime
+from typing import Optional
 
 import pandas as pd
 import requests
-from typing import Optional
 
 from fennel.datasets import dataset, field
 from fennel.featuresets import feature, featureset, extractor
@@ -102,25 +102,21 @@ def test_dag_resolution(self, client):
         assert response.status_code == requests.codes.OK, response.json()
         # /docsnip
 
-        # docsnip extract_features_api
-        feature_df = client.extract_features(
-            output_feature_list=[
-                UserFeatures,
-            ],
-            input_feature_list=[UserFeatures.userid],
+        # docsnip extract_api
+        feature_df = client.extract(
+            outputs=[UserFeatures],
+            inputs=[UserFeatures.userid],
             input_dataframe=pd.DataFrame(
                 {"UserFeatures.userid": [18232, 18234]}
             ),
         )
         self.assertEqual(feature_df.shape, (2, 7))
         # /docsnip
 
-        # docsnip extract_historical_features_api
-        response = client.extract_historical_features(
-            output_feature_list=[
-                UserFeatures,
-            ],
-            input_feature_list=[UserFeatures.userid],
+        # docsnip extract_historical_api
+        response = client.extract_historical(
+            outputs=[UserFeatures],
+            inputs=[UserFeatures.userid],
             format="pandas",
             input_dataframe=pd.DataFrame(
                 {"UserFeatures.userid": [18232, 18234], "timestamp": [now, now]}
@@ -130,7 +126,7 @@ def test_dag_resolution(self, client):
         # /docsnip
 
         with self.assertRaises(NotImplementedError) as e:
-            # docsnip extract_historical_features_s3
+            # docsnip extract_historical_s3
             from fennel.sources import S3
 
             s3 = S3(
@@ -143,11 +139,9 @@ def test_dag_resolution(self, client):
             )
             s3_output_connection = s3.bucket("bucket", prefix="output")
 
-            response = client.extract_historical_features(
-                output_feature_list=[
-                    UserFeatures,
-                ],
-                input_feature_list=[UserFeatures.userid],
+            response = client.extract_historical(
+                outputs=[UserFeatures],
+                inputs=[UserFeatures.userid],
                 format="csv",
                 timestamp_column="timestamp",
                 input_s3=s3_input_connection,

docs/examples/api-reference/rest-api.py

@@ -45,17 +45,17 @@ def test_log(self, mock_post):
         assert response.status_code == requests.codes.OK, response.json()
         # /docsnip
 
-        # docsnip rest_extract_features_api
-        url = "{}/api/v1/extract_features".format(SERVER)
+        # docsnip rest_extract_api
+        url = "{}/api/v1/extract".format(SERVER)
         headers = {"Content-Type": "application/json"}
         data = [
             {"UserFeatures.userid": 1},
             {"UserFeatures.userid": 2},
             {"UserFeatures.userid": 3},
         ]
         req = {
-            "output_features": ["UserFeatures"],
-            "input_features": ["UserFeatures.userid"],
+            "outputs": ["UserFeatures"],
+            "inputs": ["UserFeatures.userid"],
             "data": data,
             "log": True,
             "workflow": "test",
@@ -65,13 +65,13 @@ def test_log(self, mock_post):
         assert response.status_code == requests.codes.OK, response.json()
         # /docsnip
 
-        # docsnip rest_extract_features_api_columnar
-        url = "{}/api/v1/extract_features".format(SERVER)
+        # docsnip rest_extract_api_columnar
+        url = "{}/api/v1/extract".format(SERVER)
         headers = {"Content-Type": "application/json"}
         data = {"UserFeatures.userid": [1, 2, 3]}
         req = {
-            "output_features": ["UserFeatures"],
-            "input_features": ["UserFeatures.userid"],
+            "outputs": ["UserFeatures"],
+            "inputs": ["UserFeatures.userid"],
             "data": data,
             "log": True,
             "workflow": "test",

docs/examples/datasets/lookups.py

@@ -61,9 +61,9 @@ def test_user_dataset_lookup(client):
 
     res = client.log("fennel_webhook", "User", df)
     assert res.status_code == 200, res.json()
-    feature_df = client.extract_features(
-        output_feature_list=[UserFeature.in_home_city],
-        input_feature_list=[UserFeature.uid],
+    feature_df = client.extract(
+        outputs=[UserFeature.in_home_city],
+        inputs=[UserFeature.uid],
         input_dataframe=pd.DataFrame({"UserFeature.uid": [1, 2, 3]}),
     )
     assert feature_df["UserFeature.in_home_city"].tolist() == [

docs/examples/datasets/operators.py

@@ -237,9 +237,9 @@ def test_aggregate(client):
     three_days_ago = dt - timedelta(days=3)
     ts_series = pd.Series([dt, yes, dt, three_days_ago, yes])
     uids = pd.Series([1, 1, 2, 2, 2])
-    df = client.extract_historical_features(
-        input_feature_list=[UserAdStatsFeatures.uid],
-        output_feature_list=[UserAdStatsFeatures],
+    df = client.extract_historical(
+        inputs=[UserAdStatsFeatures.uid],
+        outputs=[UserAdStatsFeatures],
         input_dataframe=pd.DataFrame(
             {"UserAdStatsFeatures.uid": uids, "timestamps": ts_series}
         ),

docs/examples/examples/ecommerce.py

@@ -113,12 +113,12 @@ def test_feature(self, client):
         response = client.log("fennel_webhook", "Order", df)
         assert response.status_code == requests.codes.OK, response.json()
 
-        feature_df = client.extract_features(
-            output_feature_list=[
+        feature_df = client.extract(
+            outputs=[
                 "UserSeller.num_orders_1d",
                 "UserSeller.num_orders_1w",
             ],
-            input_feature_list=[
+            inputs=[
                 "UserSeller.uid",
                 "UserSeller.seller_id",
             ],

docs/examples/featuresets/e2e_extraction.py

@@ -51,14 +51,14 @@ class Request:
 def test_e2e_extraction(client):
     client.sync(featuresets=[User, UserPost, Request])
     # docsnip e2e_extraction
-    feature_df = client.extract_features(
-        output_feature_list=[
+    feature_df = client.extract(
+        outputs=[
             "User.age",
             "UserPost.score",
             "UserPost.affinity"
             # there are 10 features in this list
         ],
-        input_feature_list=[
+        inputs=[
             "User.id",
             "UserPost.uid",
             "UserPost.pid",

docs/examples/featuresets/overview.py

@@ -220,9 +220,9 @@ def test_multiple_features_extracted(client):
     res = client.log("fennel_webhook", "UserInfo", df)
     assert res.status_code == 200
 
-    df = client.extract_features(
-        output_feature_list=[UserLocationFeatures],
-        input_feature_list=[UserLocationFeatures.uid],
+    df = client.extract(
+        outputs=[UserLocationFeatures],
+        inputs=[UserLocationFeatures.uid],
         input_dataframe=pd.DataFrame(
             {"UserLocationFeatures.uid": [1, 2, 3]},
         ),
@@ -292,9 +292,9 @@ def test_extractors_across_featuresets(client):
     res = client.log("fennel_webhook", "UserInfo", df)
     assert res.status_code == 200
 
-    df = client.extract_features(
-        output_feature_list=[UserLocationFeaturesRefactored],
-        input_feature_list=[Request.uid],
+    df = client.extract(
+        outputs=[UserLocationFeaturesRefactored],
+        inputs=[Request.uid],
         input_dataframe=pd.DataFrame(
             {"Request.uid": [1, 2, 3]},
         ),

docs/examples/featuresets/reading_datasets.py

@@ -92,9 +92,9 @@ def test_lookup_in_extractor(client):
     res = client.log("fennel_webhook", "User", data)
     assert res.status_code == 200, res.json()
 
-    feature_df = client.extract_features(
-        output_feature_list=[UserFeatures.name],
-        input_feature_list=[UserFeatures.uid],
+    feature_df = client.extract(
+        outputs=[UserFeatures.name],
+        inputs=[UserFeatures.uid],
         input_dataframe=pd.DataFrame(
             {
                 "UserFeatures.uid": [1, 2, 3, 4],
@@ -105,9 +105,9 @@ def test_lookup_in_extractor(client):
     expected = ["Alice", "Bob", "Charlie", "Unknown"]
     assert feature_df["UserFeatures.name"].tolist() == expected
 
-    feature_df = client.extract_features(
-        output_feature_list=[UserFeaturesDerived.name],
-        input_feature_list=[Request.user_id],
+    feature_df = client.extract(
+        outputs=[UserFeaturesDerived.name],
+        inputs=[Request.user_id],
         input_dataframe=pd.DataFrame(
             {
                 "Request.user_id": [1, 2, 3, 4],
@@ -116,9 +116,9 @@ def test_lookup_in_extractor(client):
     )
     assert feature_df["UserFeaturesDerived.name"].tolist() == expected
 
-    feature_df = client.extract_features(
-        output_feature_list=[UserFeaturesDerived2.name],
-        input_feature_list=[Request2.uid],
+    feature_df = client.extract(
+        outputs=[UserFeaturesDerived2.name],
+        inputs=[Request2.uid],
         input_dataframe=pd.DataFrame(
             {
                 "Request2.uid": [1, 2, 3, 4],

docs/examples/getting-started/quickstart.py

@@ -1,9 +1,9 @@
 # docsnip imports
 from datetime import datetime, timedelta
+from typing import Optional
 
 import pandas as pd
 import requests
-from typing import Optional
 
 from fennel.datasets import dataset, pipeline, field, Dataset
 from fennel.featuresets import feature, featureset, extractor
@@ -145,12 +145,12 @@ def myextractor(cls, ts: pd.Series, uids: pd.Series, sellers: pd.Series):
 # /docsnip
 
 # docsnip query
-feature_df = client.extract_features(
-    output_feature_list=[
+feature_df = client.extract(
+    outputs=[
         UserSellerFeatures.num_orders_1d,
         UserSellerFeatures.num_orders_1w,
     ],
-    input_feature_list=[
+    inputs=[
         UserSellerFeatures.uid,
         UserSellerFeatures.seller_id,
     ],
@@ -170,12 +170,12 @@ def myextractor(cls, ts: pd.Series, uids: pd.Series, sellers: pd.Series):
 # /docsnip
 
 # docsnip historical
-feature_df = client.extract_historical_features(
-    output_feature_list=[
+feature_df = client.extract_historical(
+    outputs=[
         UserSellerFeatures.num_orders_1d,
         UserSellerFeatures.num_orders_1w,
     ],
-    input_feature_list=[
+    inputs=[
         UserSellerFeatures.uid,
         UserSellerFeatures.seller_id,
     ],

docs/examples/testing-and-ci-cd/ci_cd/test.py

@@ -1,9 +1,11 @@
 from datetime import datetime
+
 import pandas as pd
-from fennel.test_lib import mock
 from ci_cd.datasets import Ticket
 from ci_cd.featuresets import TicketFeatures
 
+from fennel.test_lib import mock
+
 
 @mock
 def test_featureset_metaflags(client):
@@ -17,9 +19,9 @@ def test_featureset_metaflags(client):
         columns=["ticket_id", "price", "at"],
     )
     client.log("example", "ticket_sale", df)
-    feature_df = client.extract_features(
-        input_feature_list=[TicketFeatures.ticket_id],
-        output_feature_list=[TicketFeatures.price, TicketFeatures.ticket_id],
+    feature_df = client.extract(
+        inputs=[TicketFeatures.ticket_id],
+        outputs=[TicketFeatures.price, TicketFeatures.ticket_id],
         input_dataframe=pd.DataFrame(
             data={"TicketFeatures.ticket_id": ["123", "456"]}
         ),

docs/examples/testing-and-ci-cd/unit_tests.py

@@ -1,8 +1,8 @@
 from datetime import datetime, timedelta
+from typing import Optional
 
 import pandas as pd
 import requests
-from typing import Optional
 
 # docsnip datasets
 from fennel.datasets import dataset, field, pipeline, Dataset
@@ -234,11 +234,9 @@ def test_dag_resolution(self, client):
         response = client.log("fennel_webhook", "UserInfoDataset", df)
         assert response.status_code == requests.codes.OK, response.json()
 
-        feature_df = client.extract_features(
-            output_feature_list=[
-                UserInfoMultipleExtractor,
-            ],
-            input_feature_list=[UserInfoMultipleExtractor.userid],
+        feature_df = client.extract(
+            outputs=[UserInfoMultipleExtractor],
+            inputs=[UserInfoMultipleExtractor.userid],
             input_dataframe=pd.DataFrame(
                 {"UserInfoMultipleExtractor.userid": [18232, 18234]}
             ),