[Docs] Change feature store alias to fstore (from fs) (#1001)

docs/feature-store/basic-demo.ipynb

@@ -434,7 +434,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import mlrun.feature_store as fs\n",
+    "import mlrun.feature_store as fstore\n",
     "from mlrun.feature_store.steps import *\n",
     "from mlrun.features import MinMaxValidator"
    ]
@@ -516,8 +516,8 @@
    ],
    "source": [
     "# add feature set without time column (stock ticker metadata) \n",
-    "stocks_set = fs.FeatureSet(\"stocks\", entities=[fs.Entity(\"ticker\")])\n",
-    "fs.ingest(stocks_set, stocks, infer_options=fs.InferOptions.default())"
+    "stocks_set = fstore.FeatureSet(\"stocks\", entities=[fstore.Entity(\"ticker\")])\n",
+    "fstore.ingest(stocks_set, stocks, infer_options=fstore.InferOptions.default())"
    ]
   },
   {
@@ -535,7 +535,7 @@
    "outputs": [],
    "source": [
     "# create a new feature set\n",
-    "quotes_set = fs.FeatureSet(\"stock-quotes\", entities=[fs.Entity(\"ticker\")])"
+    "quotes_set = fstore.FeatureSet(\"stock-quotes\", entities=[fstore.Entity(\"ticker\")])"
    ]
   },
   {
@@ -704,7 +704,7 @@
     "quotes_set.add_aggregation(\"bids\", \"bid\", [\"min\", \"max\"], \"1h\", \"10m\")\n",
     "\n",
     "# add feature validation policy\n",
-    "quotes_set[\"bid\"] = fs.Feature(validator=MinMaxValidator(min=52, severity=\"info\"))\n",
+    "quotes_set[\"bid\"] = fstore.Feature(validator=MinMaxValidator(min=52, severity=\"info\"))\n",
     "\n",
     "# add default target definitions and plot\n",
     "quotes_set.set_targets()\n",
@@ -938,12 +938,12 @@
     }
    ],
    "source": [
-    "fs.infer_metadata(\n",
+    "fstore.infer_metadata(\n",
     "    quotes_set,\n",
     "    quotes,\n",
     "    entity_columns=[\"ticker\"],\n",
     "    timestamp_key=\"time\",\n",
-    "    options=fs.InferOptions.default(),\n",
+    "    options=fstore.InferOptions.default(),\n",
     ")"
    ]
   },
@@ -1699,7 +1699,7 @@
    ],
    "source": [
     "# save ingest data and print the FeatureSet spec\n",
-    "df = fs.ingest(quotes_set, quotes)"
+    "df = fstore.ingest(quotes_set, quotes)"
    ]
   },
   {
@@ -1726,7 +1726,7 @@
     "    \"stocks.*\",\n",
     "]\n",
     "\n",
-    "vector = fs.FeatureVector(\"stocks-vec\", features, description=\"stocks demo feature vector\")\n",
+    "vector = fstore.FeatureVector(\"stocks-vec\", features, description=\"stocks demo feature vector\")\n",
     "vector.save()"
    ]
   },
@@ -1848,7 +1848,7 @@
     }
    ],
    "source": [
-    "resp = fs.get_offline_features(vector, entity_rows=trades, entity_timestamp_column=\"time\")\n",
+    "resp = fstore.get_offline_features(vector, entity_rows=trades, entity_timestamp_column=\"time\")\n",
     "resp.to_dataframe()"
    ]
   },
@@ -1865,7 +1865,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "service = fs.get_online_feature_service(\"stocks-vec\")"
+    "service = fstore.get_online_feature_service(\"stocks-vec\")"
    ]
   },
   {

docs/feature-store/end-to-end-demo/01-ingest-datasources.ipynb

@@ -37,21 +37,16 @@
    "outputs": [],
    "source": [
     "import mlrun\n",
-    "from os import getenv\n",
     "\n",
-    "mlrun.set_environment(project='fsdemo', user_project=True)\n",
+    "project, artifact_path = mlrun.set_environment(project='fsdemo', user_project=True)\n",
     "# location of the output data files\n",
-    "data_path = f\"{getenv('V3IO_HOME_URL')}/demos/feature-store/data/\""
+    "data_path = f\"{artifact_path}/data/\""
    ]
   },
   {
    "cell_type": "code",
    "execution_count": 2,
-   "metadata": {
-    "jupyter": {
-     "source_hidden": true
-    }
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "def move_timestamps(df, shift='0s'):\n",
@@ -86,7 +81,7 @@
    "outputs": [],
    "source": [
     "# Import MLRun's Feature Store\n",
-    "import mlrun.feature_store as fs\n",
+    "import mlrun.feature_store as fstore\n",
     "\n",
     "# Import MLRun's Data Sources to set the wanted ingestion pipeline\n",
     "from mlrun.datastore.sources import CSVSource, ParquetSource, HttpSource"
@@ -103,8 +98,8 @@
     "import storey\n",
     "\n",
     "# Define the Lab Measurements FeatureSet\n",
-    "measurements_set = fs.FeatureSet(\"measurements\", \n",
-    "                                 entities=[fs.Entity(\"patient_id\")], \n",
+    "measurements_set = fstore.FeatureSet(\"measurements\", \n",
+    "                                 entities=[fstore.Entity(\"patient_id\")], \n",
     "                                 timestamp_key='timestamp', \n",
     "                                 description=\"various patient health measurements\")\n",
     "\n",
@@ -676,8 +671,8 @@
    "source": [
     "# User our loaded DF as the datasource and ingest it through\n",
     "# the define pipeline\n",
-    "resp = fs.ingest(measurements_set, measurements_df, \n",
-    "                 infer_options=fs.InferOptions.default())\n",
+    "resp = fstore.ingest(measurements_set, measurements_df, \n",
+    "                     infer_options=fstore.InferOptions.default())\n",
     "resp.head()"
    ]
   },
@@ -720,7 +715,7 @@
    "outputs": [],
    "source": [
     "# add feature set without time column (stock ticker metadata) \n",
-    "patients_set = fs.FeatureSet(\"patient_details\", entities=[fs.Entity(\"patient_id\")],\n",
+    "patients_set = fstore.FeatureSet(\"patient_details\", entities=[fstore.Entity(\"patient_id\")],\n",
     "                             description=\"personal and medical patient details\")\n",
     "\n",
     "# Get FeatureSet computation graph\n",
@@ -1018,7 +1013,7 @@
     "patients_df = pd.read_parquet('https://s3.wasabisys.com/iguazio/data/patients/patient_details.parquet')\n",
     "\n",
     "# Run local ingestion test\n",
-    "fs.infer(patients_set, patients_df.head())"
+    "fstore.infer(patients_set, patients_df.head())"
    ]
   },
   {
@@ -1040,8 +1035,8 @@
     "patients_set.save()\n",
     "\n",
     "# Run Ingestion task\n",
-    "resp = fs.ingest(patients_set, patients_df, \n",
-    "                 infer_options=fs.InferOptions.default())"
+    "resp = fstore.ingest(patients_set, patients_df, \n",
+    "                     infer_options=fstore.InferOptions.default())"
    ]
   },
   {
@@ -1303,8 +1298,8 @@
    ],
    "source": [
     "source = ParquetSource('pq', 'https://s3.wasabisys.com/iguazio/data/patients/patient_details.parquet')\n",
-    "config = fs.RunConfig(local=False).apply(mlrun.platforms.auto_mount())\n",
-    "fs.ingest(patients_set, source, run_config=config)"
+    "config = fstore.RunConfig(local=False).apply(mlrun.platforms.auto_mount())\n",
+    "fstore.ingest(patients_set, source, run_config=config)"
    ]
   },
   {
@@ -1329,8 +1324,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "early_sense_set = fs.FeatureSet(\"early_sense\", entities=[fs.Entity(\"patient_id\")], timestamp_key='timestamp',\n",
-    "                                description=\"real time patient bed sensor data\")"
+    "early_sense_set = fstore.FeatureSet(\"early_sense\", entities=[fstore.Entity(\"patient_id\")], timestamp_key='timestamp',\n",
+    "                                    description=\"real time patient bed sensor data\")"
    ]
   },
   {
@@ -1350,8 +1345,8 @@
    "source": [
     "from mlrun.features import MinMaxValidator\n",
     "\n",
-    "early_sense_set[\"hr\"] = fs.Feature(validator = MinMaxValidator(min=0, max=220, severity=\"info\"))\n",
-    "early_sense_set[\"rr\"] = fs.Feature(validator = MinMaxValidator(min=0, max=25, severity=\"info\"))"
+    "early_sense_set[\"hr\"] = fstore.Feature(validator = MinMaxValidator(min=0, max=220, severity=\"info\"))\n",
+    "early_sense_set[\"rr\"] = fstore.Feature(validator = MinMaxValidator(min=0, max=25, severity=\"info\"))"
    ]
   },
   {
@@ -1829,7 +1824,7 @@
     "early_sense_df = pd.read_parquet('https://s3.wasabisys.com/iguazio/data/patients/early_sense.parquet')\n",
     "early_sense_df['timestamp'] = pd.to_datetime(early_sense_df['timestamp'])\n",
     "early_sense_df = move_timestamps(early_sense_df) # update timestamps\n",
-    "fs.infer(early_sense_set, early_sense_df.head())"
+    "fstore.infer(early_sense_set, early_sense_df.head())"
    ]
   },
   {
@@ -1847,7 +1842,7 @@
    ],
    "source": [
     "# Run ingest pipeline\n",
-    "df=fs.ingest(early_sense_set, early_sense_df)\n",
+    "df=fstore.ingest(early_sense_set, early_sense_df)\n",
     "\n",
     "# Save the early-sense Featureset\n",
     "early_sense_set.save()\n",
@@ -1909,11 +1904,11 @@
     "# an MLRun function from them so we can run the pipeline\n",
     "# as a Nuclio function\n",
     "func = mlrun.code_to_function(\"ingest\", kind=\"serving\")\n",
-    "nuclio_config = fs.RunConfig(function=func, local=False).apply(mlrun.platforms.auto_mount())\n",
+    "nuclio_config = fstore.RunConfig(function=func, local=False).apply(mlrun.platforms.auto_mount())\n",
     "\n",
     "# Deploy the Online ingestion service using the pipeline definition from before\n",
     "# with our new HTTP Source and our define Function\n",
-    "server = fs.deploy_ingestion_service(early_sense_set, source, run_config=nuclio_config)"
+    "server = fstore.deploy_ingestion_service(early_sense_set, source, run_config=nuclio_config)"
    ]
   },
   {
@@ -2024,10 +2019,10 @@
    ],
    "source": [
     "#labels_df = pd.read_parquet('labels.parquet')\n",
-    "labels_set = fs.FeatureSet(\"labels\", entities=[fs.Entity(\"patient_id\")], timestamp_key='timestamp',\n",
-    "                           description=\"training labels\")\n",
+    "labels_set = fstore.FeatureSet(\"labels\", entities=[fstore.Entity(\"patient_id\")], timestamp_key='timestamp',\n",
+    "                               description=\"training labels\")\n",
     "labels_set.set_targets()\n",
-    "df = fs.infer(labels_set, data_path + 'labels.parquet')"
+    "df = fstore.infer(labels_set, data_path + 'labels.parquet')"
    ]
   },
   {
@@ -2178,7 +2173,7 @@
     }
    ],
    "source": [
-    "df = fs.ingest(labels_set, data_path + 'labels.parquet')"
+    "df = fstore.ingest(labels_set, data_path + 'labels.parquet')"
    ]
   },
   {
@@ -2221,4 +2216,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

docs/feature-store/end-to-end-demo/02-create-training-model.ipynb

@@ -31,11 +31,10 @@
    "outputs": [],
    "source": [
     "import mlrun\n",
-    "from os import getenv\n",
     "\n",
-    "project, _ = mlrun.set_environment(project='fsdemo', user_project=True)\n",
+    "project, artifact_path = mlrun.set_environment(project='fsdemo', user_project=True)\n",
     "# location of the output data files\n",
-    "data_path = f\"{getenv('V3IO_HOME_URL')}/demos/feature-store/data/\""
+    "data_path = f\"{artifact_path}/data/\""
    ]
   },
   {
@@ -58,7 +57,7 @@
    "outputs": [],
    "source": [
     "# Import MLRun's Feature Store\n",
-    "import mlrun.feature_store as fs\n",
+    "import mlrun.feature_store as fstore\n",
     "\n",
     "# Define the featuer vector's name for future reference\n",
     "feature_vector_name = 'patient-deterioration'\n",
@@ -101,10 +100,10 @@
     "            ]\n",
     "\n",
     "# Define the feature vector\n",
-    "fv = fs.FeatureVector(feature_vector_name, \n",
-    "                      features, \n",
-    "                      label_feature=\"labels.label\",\n",
-    "                      description='Predict patient deterioration')\n",
+    "fv = fstore.FeatureVector(feature_vector_name, \n",
+    "                          features, \n",
+    "                          label_feature=\"labels.label\",\n",
+    "                          description='Predict patient deterioration')\n",
     "\n",
     "# Save the feature vector in the Feature Store\n",
     "fv.save()"
@@ -350,7 +349,7 @@
     "# Get offline feature vector\n",
     "# will return a pandas dataframe and save the dataset to parquet so a \n",
     "# training job could train on it\n",
-    "dataset = fs.get_offline_features(feature_vector_name, target=ParquetTarget())\n",
+    "dataset = fstore.get_offline_features(feature_vector_name, target=ParquetTarget())\n",
     "\n",
     "# View dataset example\n",
     "df = dataset.to_dataframe()\n",
@@ -922,7 +921,7 @@
     }
    ],
    "source": [
-    "fs.get_feature_vector(f'{project}/{feature_vector_name}').status.targets['parquet'].to_dict()"
+    "fstore.get_feature_vector(f'{project}/{feature_vector_name}').status.targets['parquet'].to_dict()"
    ]
   },
   {

docs/feature-store/feature-sets.md

@@ -57,13 +57,14 @@ in a NoSQL DB, users can use the default targets or add/replace with additional
 
 Graph example (storey engine):
 ```python
-feature_set = FeatureSet("measurements", entities=[Entity(key)], timestamp_key="timestamp")
+import mlrun.feature_store as fstore
+feature_set = fstore.FeatureSet("measurements", entities=[Entity(key)], timestamp_key="timestamp")
 # Define the computational graph including our custom functions
 feature_set.graph.to(DropColumns(drop_columns))\
                  .to(RenameColumns(mapping={'bad': 'bed'}))
 feature_set.add_aggregation('hr', 'hr', ['avg'], ["1h"])
 feature_set.plot()
-fs.ingest(feature_set, data_df)
+fstore.ingest(feature_set, data_df)
 ```
 
 Graph example (pandas engine):
@@ -72,9 +73,9 @@ def myfunc1(df, context=None):
     df = df.drop(columns=["exchange"])
     return df
 
-stocks_set = fs.FeatureSet("stocks", entities=[Entity("ticker")], engine="pandas")
+stocks_set = fstore.FeatureSet("stocks", entities=[Entity("ticker")], engine="pandas")
 stocks_set.graph.to(name="s1", handler="myfunc1")
-df = fs.ingest(stocks_set, stocks_df)
+df = fstore.ingest(stocks_set, stocks_df)
 ```
 
 The graph steps can use built-in transformation classes, simple python classes or function handlers. 
@@ -85,7 +86,7 @@ This allows to get a preview of the results (in the returned dataframe). The sim
 The infer operation also learns the feature set schema and does statistical analysis on the result by default.
 
 ```python
-df = fs.infer(quotes_set, quotes)
+df = fstore.infer(quotes_set, quotes)
 
 # print the featue statistics
 print(quotes_set.get_stats_table())
@@ -146,7 +147,7 @@ There are multiple data source options including http, kafka, kinesis, v3io stre
 source = HTTPSource()
 func = mlrun.code_to_function("ingest", kind="serving").apply(mount_v3io())
 config = RunConfig(function=func)
-fs.deploy_ingestion_service(my_set, source, run_config=config)
+fstore.deploy_ingestion_service(my_set, source, run_config=config)
 ```
 
 To learn more about deploy_ingestion_service go to {py:class}`~mlrun.feature_store.deploy_ingestion_service`