Capture dataset-related events (#385)

* add events to dataset activities

* properly raise exception in listing dataset history call

* record events on dataset testing

kolena/_api/v1/event.py

@@ -54,6 +54,15 @@ class Event(str, Enum):
         # test run
         EXECUTE_TEST_RUN = "sdk-test-run-executed"
 
+        # dataset
+        REGISTER_DATASET = "sdk-dataset-registered"
+        FETCH_DATASET = "sdk-dataset-fetched"
+        FETCH_DATASET_HISTORY = "sdk-dataset-history-fetched"
+
+        # dataset evaluation
+        FETCH_DATASET_MODEL_RESULT = "sdk-dataset-model-result-fetched"
+        TEST_DATASET_MODEL = "sdk-dataset-model-tested"
+
     @dataclass(frozen=True)
     class RecordEventRequest:
         event_name: str

kolena/dataset/dataset.py

@@ -25,6 +25,7 @@
 import pandas as pd
 import requests
 
+from kolena._api.v1.event import EventAPI
 from kolena._api.v2.dataset import CommitData
 from kolena._api.v2.dataset import EntityData
 from kolena._api.v2.dataset import ListCommitHistoryRequest
@@ -40,6 +41,7 @@
 from kolena._utils.batched_load import upload_data_frame
 from kolena._utils.consts import BatchSize
 from kolena._utils.endpoints import get_dataset_url
+from kolena._utils.instrumentation import with_event
 from kolena._utils.serde import from_dict
 from kolena._utils.state import API_V2
 from kolena.dataset.common import COL_DATAPOINT
@@ -227,6 +229,7 @@ def resolve_id_fields(
     return id_fields
 
 
+@with_event(event_name=EventAPI.Event.REGISTER_DATASET)
 def register_dataset(
     name: str,
     df: Union[Iterator[pd.DataFrame], pd.DataFrame],
@@ -296,6 +299,7 @@ def _iter_dataset(
         yield _to_deserialized_dataframe(df_batch, column=COL_DATAPOINT)
 
 
+@with_event(event_name=EventAPI.Event.FETCH_DATASET)
 def fetch_dataset(
     name: str,
     commit: str = None,
@@ -315,7 +319,7 @@ def _list_commits(name: str, descending: bool = False, offset: int = 0, limit: i
     """
     request = ListCommitHistoryRequest(name=name, descending=descending, offset=offset, limit=limit)
     response = krequests.put(Path.LIST_COMMITS, json=asdict(request))
-    response.raise_for_status()
+    krequests.raise_for_status(response)
     return from_dict(ListCommitHistoryResponse, response.json())
 
 
@@ -339,6 +343,7 @@ def _iter_commits(
             break
 
 
+@with_event(event_name=EventAPI.Event.FETCH_DATASET_HISTORY)
 def fetch_dataset_history(
     name: str,
     descending: bool = False,

kolena/dataset/evaluation.py

@@ -23,6 +23,7 @@
 
 import pandas as pd
 
+from kolena._api.v1.event import EventAPI
 from kolena._api.v2.model import LoadResultsRequest
 from kolena._api.v2.model import Path
 from kolena._api.v2.model import UploadResultsRequest
@@ -32,6 +33,7 @@
 from kolena._utils.batched_load import init_upload
 from kolena._utils.batched_load import upload_data_frame
 from kolena._utils.consts import BatchSize
+from kolena._utils.instrumentation import with_event
 from kolena._utils.state import API_V2
 from kolena.dataset.common import COL_DATAPOINT
 from kolena.dataset.common import COL_DATAPOINT_ID_OBJECT
@@ -100,6 +102,7 @@ def _upload_results(
     krequests.raise_for_status(response)
 
 
+@with_event(EventAPI.Event.FETCH_DATASET_MODEL_RESULT)
 def fetch_results(
     dataset: str,
     model: str,
@@ -133,6 +136,7 @@ def _validate_configs(configs: List[TYPE_EVALUATION_CONFIG]) -> None:
                 raise IncorrectUsageError("duplicate eval configs are invalid")
 
 
+@with_event(EventAPI.Event.TEST_DATASET_MODEL)
 def test(
     dataset: str,
     model: str,