remove usage stats and various cleanup (#1596)

## Description
Remove the usage_stats emit_usage telemetry from whylogs and various
cleanup:
* remove whylabs API key format length checks
* more verbose retry warnings in failures using the whylabs_client
* test cleanup

- [x] I have reviewed the [Guidelines for Contributing](CONTRIBUTING.md)
and the [Code of Conduct](CODE_OF_CONDUCT.md).

Author: jamie256

python/tests/api/writer/test_whylabs.py

@@ -342,22 +342,6 @@ def test_option_will_overwrite_defaults(self) -> None:
         assert writer._whylabs_client._dataset_id == "new_dataset_id"
         assert writer.key_id == "newkeynewk"
 
-    def test_api_key_prefers_parameter_over_env_var(self, results, caplog):
-        with pytest.raises(ValueError):
-            results.writer("whylabs").option(org_id="org_id", api_key="api_key_123.foo").write(dataset_id="dataset_id")
-
-    def test_writer_accepts_dest_param(self, results, caplog):
-        # TODO: inspect error or mock better to avoid network call and keep test focused.
-        with pytest.raises(ValueError):
-            results.writer("whylabs").option(api_key="bad_key_format").write(dataset_id="dataset_id", dest="tmp")
-
-    def test_write_response(self, results):
-        with pytest.raises(ValueError):
-            response = (
-                results.writer("whylabs").option(api_key="bad_key_format").write(dataset_id="dataset_id", dest="tmp")
-            )
-            assert response[0] is True
-
     def test_changing_api_key_works(self) -> None:
         #
         # Defaults

python/whylogs/__init__.py

@@ -26,7 +26,6 @@
     reader,
     write,
 )
-from .api.usage_stats import emit_usage as __emit_usage_stats
 from .api.whylabs import init
 from .core import DatasetProfileView
 from .migration.converters import v0_to_v1_view
@@ -65,5 +64,3 @@ def package_version(package: str = __package__) -> str:
     __version__,
     init,
 ]
-
-__emit_usage_stats("import")

python/whylogs/api/fugue/__init__.py

@@ -1,7 +1,3 @@
-# flake8: noqa
-from whylogs.api.usage_stats import emit_usage
+from .profiler import fugue_profile
 
-# This import has a side effect
-from .profiler import fugue_profile  # type: ignore
-
-emit_usage("fugue")
+assert fugue_profile is not None

python/whylogs/api/logger/__init__.py

@@ -20,7 +20,6 @@
     _log_segment,
 )
 from whylogs.api.logger.transient import TransientLogger
-from whylogs.api.usage_stats import emit_usage
 from whylogs.api.whylabs.session.notebook_logger import (
     notebook_session_log,
     notebook_session_log_comparison,
@@ -55,7 +54,6 @@ def log(
 ) -> ResultSet:
     if multiple is not None:
         result_sets: Dict[str, ResultSet] = {}
-        emit_usage("multiple")
         for alias, data in multiple.items():
             result_set = TransientLogger(schema=schema).log(data, trace_id=trace_id)
             if dataset_timestamp is not None:

python/whylogs/api/pyspark/experimental/profiler.py

@@ -4,15 +4,13 @@
 from typing import Dict, Iterable, Optional, Tuple
 
 import whylogs as why
-from whylogs.api.usage_stats import emit_usage
 from whylogs.core import DatasetSchema
 from whylogs.core.metrics.metrics import conf
 from whylogs.core.stubs import pd
 from whylogs.core.view.column_profile_view import ColumnProfileView
 from whylogs.core.view.dataset_profile_view import DatasetProfileView
 
 logger = getLogger(__name__)
-emit_usage("pyspark")
 
 try:  # type: ignore
     from pyspark.ml.functions import vector_to_array

python/whylogs/api/pyspark/experimental/segmented_profiler.py

@@ -8,7 +8,6 @@
 import whylogs as why
 from whylogs.api.logger.result_set import ResultSet, SegmentedResultSet
 from whylogs.api.pyspark.experimental.profiler import COL_NAME_FIELD, COL_PROFILE_FIELD
-from whylogs.api.usage_stats import emit_usage
 from whylogs.core import DatasetSchema
 from whylogs.core.segment import Segment
 from whylogs.core.segmentation_partition import SegmentationPartition
@@ -17,7 +16,6 @@
 from whylogs.core.view.dataset_profile_view import DatasetProfileView
 
 logger = getLogger(__name__)
-emit_usage("pyspark")
 
 try:  # type: ignore
     from pyspark.ml.functions import vector_to_array

python/whylogs/api/usage_stats/__init__.py

@@ -77,14 +77,12 @@ def __init__(self, config: SessionConfig) -> None:
         If neither exist then this will attempt to create a new session and store the id in the config,
         which does require a successful service call to whylabs.
         """
-        from whylogs.api.usage_stats import emit_usage
 
         super().__init__(config)
 
         # Using lazy initialization to work around circular dependency issues
         self._whylabs_session_api = Lazy(self.__create_session_api)
         self._user_guid = self._get_or_create_user_guid()
-        emit_usage("guest_session")
 
     def __create_session_api(self) -> SessionsApi:
         from whylogs.api.whylabs.session.whylabs_client_cache import ClientCacheConfig
@@ -255,15 +253,12 @@ def upload_batch_profile(self, profile: ResultSet) -> Union[UploadResult, NotSup
 
 class ApiKeySession(Session):
     def __init__(self, config: SessionConfig) -> None:
-        from whylogs.api.usage_stats import emit_usage
-
         super().__init__(config)
         self.api_key = config.get_api_key()
         self.org_id = config.get_org_id()
 
         # Using lazy initialization to work around circular dependency issues
         self._whylabs_log_api = Lazy(partial(self.__create_log_api, config))
-        emit_usage("api_key_session")
 
     def __create_log_api(self, config: SessionConfig) -> LogApi:
         from whylogs.api.whylabs.session.whylabs_client_cache import ClientCacheConfig

python/whylogs/api/whylabs/session/session.py

@@ -22,10 +22,6 @@ def __call__(self, config: Configuration) -> None:
     def _validate_api_key(self, api_key: Optional[str]) -> str:
         if api_key is None:
             raise ValueError("Missing API key. Set it via WHYLABS_API_KEY environment variable or as an api_key option")
-        if len(api_key) < 12:
-            raise ValueError("API key too short")
-        if api_key[10] != ".":
-            raise ValueError("Invalid format. Expecting a dot at an index 10")
         return api_key[:10]
 
 

python/whylogs/api/whylabs/session/whylabs_client_cache.py

@@ -5,7 +5,6 @@
 
 import mlflow
 
-from whylogs.api.usage_stats import emit_usage
 from whylogs.api.writer import Writer
 from whylogs.api.writer.writer import _Writable
 from whylogs.core.utils import deprecated_alias
@@ -18,7 +17,6 @@ def __init__(self) -> None:
         self._file_dir = "whylogs"
         self._file_name = None
         self._end_run = True
-        emit_usage("mlflow_writer")
 
     @deprecated_alias(profile="file")
     def write(