arize 8.0.0a23__py3-none-any.whl → 8.0.0b0__py3-none-any.whl

arize/__init__.py

@@ -44,18 +44,19 @@ def make_to_df(field_name: str) -> object:
           - If an item is a mapping (dict-like), use it as-is.
           - Otherwise, raise a ValueError (unsupported row type).
 
-        Parameters:
-          by_alias: Use field aliases when dumping Pydantic models.
-          exclude_none:
-            - False: keep Nones as-is
-            - "all": drop columns where *all* values are None/NaN
-            - "any": drop columns where *any* value is None/NaN
-            - True: alias for "all"
-          json_normalize: If True, flatten nested dicts via `pandas.json_normalize`.
-          convert_dtypes: If True, call `DataFrame.convert_dtypes()` at the end.
+        Args:
+            self (object): The object instance containing the field to convert.
+            by_alias (bool): Use field aliases when dumping Pydantic models.
+            exclude_none (str | bool): Control None/NaN column dropping.
+                - False: keep Nones as-is
+                - "all": drop columns where all values are None/NaN
+                - "any": drop columns where any value is None/NaN
+                - True: alias for "all"
+            json_normalize (bool): If True, flatten nested dicts via `pandas.json_normalize`.
+            convert_dtypes (bool): If True, call `DataFrame.convert_dtypes()` at the end.
 
         Returns:
-          pandas.DataFrame
+            pandas.DataFrame: The converted DataFrame.
         """
         import pandas as pd
 

arize/_exporter/client.py

@@ -17,7 +17,7 @@ from arize._exporter.validation import (
 )
 from arize._generated.protocol.flight import flight_pb2
 from arize.logging import CtxAdapter
-from arize.types import Environments, SimilaritySearchParams
+from arize.ml.types import Environments, SimilaritySearchParams
 from arize.utils.dataframe import reset_dataframe_index
 
 logger = logging.getLogger(__name__)

arize/client.py

@@ -13,13 +13,21 @@ from arize.config import SDKConfiguration
 if TYPE_CHECKING:
     from arize.datasets.client import DatasetsClient
     from arize.experiments.client import ExperimentsClient
-    from arize.models.client import MLModelsClient
+    from arize.ml.client import MLModelsClient
     from arize.projects.client import ProjectsClient
     from arize.regions import Region
     from arize.spans.client import SpansClient
 
 logger = logging.getLogger(__name__)
 
+# TODO(Kiko): Clean commented lines over the SDK
+# TODO(Kiko): Implement https://github.com/Arize-ai/arize/pull/59917
+
+# TODO(Kiko): Go private connect. Need a `base_domain`, such that we get:
+# - api.<base_domain>
+# - app.<base_domain>
+# - flight.<base_domain>
+# - otlp.<base_domain>
 
 # TODO(Kiko): Enforce type checking, remove all type ignores
 
@@ -44,133 +52,35 @@ class ArizeClient(LazySubclientsMixin):
     """Root client for the Arize SDK.
 
     The ArizeClient provides access to all Arize platform services including datasets,
-    experiments, models, projects, and spans. It uses SDKConfiguration internally to
+    experiments, ML models, projects, and spans. It uses SDKConfiguration internally to
     manage configuration settings.
 
-    Configuration Precedence
-    ------------------------
     All parameters are optional (except api_key which must be provided via argument
     or environment variable). For each parameter, values are resolved in this order:
-        1. Explicit value passed to constructor (highest priority)
-        2. Environment variable value (see SDKConfiguration for variable names)
-        3. Built-in default value (lowest priority)
-
-    Parameters
-    ----------
-    api_key : str
-        Arize API key for authentication. Required - must be provided via this parameter
-        or the ARIZE_API_KEY environment variable.
-        Environment variable: ARIZE_API_KEY
-        Default: None (raises MissingAPIKeyError if not provided)
-
-    region : Region, optional
-        Arize region (e.g., Region.US_CENTRAL, Region.EU_WEST). When specified,
-        overrides individual host/port settings with region-specific endpoints.
-        Environment variable: ARIZE_REGION
-        Default: Region.UNSPECIFIED
-
-    api_host : str, optional
-        Custom API endpoint host override.
-        Environment variable: ARIZE_API_HOST
-        Default: "api.arize.com"
-
-    api_scheme : str, optional
-        API endpoint scheme (http/https).
-        Environment variable: ARIZE_API_SCHEME
-        Default: "https"
-
-    otlp_host : str, optional
-        OTLP (OpenTelemetry Protocol) endpoint host override.
-        Environment variable: ARIZE_OTLP_HOST
-        Default: "otlp.arize.com"
-
-    otlp_scheme : str, optional
-        OTLP endpoint scheme (http/https).
-        Environment variable: ARIZE_OTLP_SCHEME
-        Default: "https"
-
-    flight_host : str, optional
-        Apache Arrow Flight endpoint host override.
-        Environment variable: ARIZE_FLIGHT_HOST
-        Default: "flight.arize.com"
-
-    flight_port : int, optional
-        Apache Arrow Flight endpoint port (1-65535).
-        Environment variable: ARIZE_FLIGHT_PORT
-        Default: 443
-
-    flight_scheme : str, optional
-        Apache Arrow Flight endpoint scheme.
-        Environment variable: ARIZE_FLIGHT_SCHEME
-        Default: "grpc+tls"
-
-    pyarrow_max_chunksize : int, optional
-        Maximum chunk size for PyArrow operations (1 to MAX_CHUNKSIZE).
-        Environment variable: ARIZE_MAX_CHUNKSIZE
-        Default: 10_000
-
-    request_verify : bool, optional
-        Whether to verify SSL certificates for HTTP requests.
-        Environment variable: ARIZE_REQUEST_VERIFY
-        Default: True
-
-    stream_max_workers : int, optional
-        Maximum number of worker threads for streaming operations (minimum: 1).
-        Environment variable: ARIZE_STREAM_MAX_WORKERS
-        Default: 8
-
-    stream_max_queue_bound : int, optional
-        Maximum queue size for streaming operations (minimum: 1).
-        Environment variable: ARIZE_STREAM_MAX_QUEUE_BOUND
-        Default: 5000
-
-    max_http_payload_size_mb : float, optional
-        Maximum HTTP payload size in megabytes (minimum: 1).
-        Environment variable: ARIZE_MAX_HTTP_PAYLOAD_SIZE_MB
-        Default: 100
-
-    arize_directory : str, optional
-        Directory for Arize SDK files (cache, logs, etc.).
-        Environment variable: ARIZE_DIRECTORY
-        Default: "~/.arize"
-
-    enable_caching : bool, optional
-        Whether to enable local caching.
-        Environment variable: ARIZE_ENABLE_CACHING
-        Default: True
-
-    single_host : str, optional
-        Single host to use for all endpoints. Overrides individual host settings.
-        Environment variable: ARIZE_SINGLE_HOST
-        Default: None (not set)
-
-    single_port : int, optional
-        Single port to use for all endpoints. Overrides individual port settings (0-65535).
-        Environment variable: ARIZE_SINGLE_PORT
-        Default: 0 (not set)
 
-    Examples:
-    --------
-    Initialize with API key only (other settings use defaults):
+    1. Explicit value passed to constructor (highest priority)
+    2. Environment variable value (see SDKConfiguration for variable names)
+    3. Built-in default value (lowest priority)
 
-        >>> client = ArizeClient(api_key="your-api-key")
+    Examples:
+        Initialize with API key only (other settings use defaults):
 
-    Initialize with custom endpoints:
+            >>> client = ArizeClient(api_key="your-api-key")
 
-        >>> client = ArizeClient(
-        ...     api_key="your-api-key",
-        ...     api_host="custom.api.com",
-        ...     flight_port=8443,
-        ... )
+        Initialize with custom endpoints:
 
-    Initialize with region (overrides host/port settings):
+            >>> client = ArizeClient(
+            ...     api_key="your-api-key",
+            ...     api_host="custom.api.com",
+            ...     flight_port=8443,
+            ... )
 
-        >>> from arize.regions import Region
-        >>> client = ArizeClient(api_key="your-api-key", region=Region.EU_WEST)
+        Initialize with region (overrides host/port settings):
 
-    See Also:
-    --------
-    SDKConfiguration : Internal configuration class with detailed parameter documentation
+            >>> from arize.regions import Region
+            >>> client = ArizeClient(
+            ...     api_key="your-api-key", region=Region.EU_WEST
+            ... )
     """
 
     _SUBCLIENTS: ClassVar[dict[str, tuple[str, str]]] = {
@@ -186,8 +96,8 @@ class ArizeClient(LazySubclientsMixin):
             "arize.projects.client",
             "ProjectsClient",
         ),
-        "models": (
-            "arize.models.client",
+        "ml": (
+            "arize.ml.client",
             "MLModelsClient",
         ),
         "spans": (
@@ -231,7 +141,7 @@ class ArizeClient(LazySubclientsMixin):
         # Imports are gated in each method of the models client
         # This is to allow for very lean package install if people only
         # want to stream ML records
-        "models": (None, ()),
+        "ml": (None, ()),
     }
 
     def __init__(
@@ -299,8 +209,8 @@ class ArizeClient(LazySubclientsMixin):
             MissingAPIKeyError: If api_key is not provided via argument or environment variable.
 
         Notes:
-            Values provided here override environment variables, which in turn override
-            default values. See SDKConfiguration for detailed parameter documentation.
+            Values provided to this class override environment variables, which in turn
+            override default values. See SDKConfiguration for detailed parameter documentation.
         """
         cfg_kwargs: dict = {}
         if api_key is not None:
@@ -356,9 +266,9 @@ class ArizeClient(LazySubclientsMixin):
         return self.__getattr__("experiments")
 
     @property
-    def models(self) -> MLModelsClient:
-        """Access the models client for ML model operations (lazy-loaded)."""
-        return self.__getattr__("models")
+    def ml(self) -> MLModelsClient:
+        """Access the ML models client for ML model operations (lazy-loaded)."""
+        return self.__getattr__("ml")
 
     @property
     def projects(self) -> ProjectsClient:
@@ -385,7 +295,7 @@ class ArizeClient(LazySubclientsMixin):
         #     'datasets': lazy,
         #     'experiments': lazy,
         #     'spans': lazy,
-        #     'models': lazy,
+        #     'ml': lazy,
         #   }
         # )
         lines = [f"{self.__class__.__name__}("]

arize/config.py

@@ -167,112 +167,71 @@ def _parse_bool(val: bool | str | None) -> bool:
 class SDKConfiguration:
     """Configuration for the Arize SDK with endpoint and authentication settings.
 
-    This class is used internally by ArizeClient to manage SDK configuration. Users
-    typically interact with ArizeClient rather than instantiating this class directly.
+    This class is used internally by ArizeClient to manage SDK configuration. It is not
+    recommended to use this class directly; users should interact with ArizeClient
+    instead.
 
-    Configuration Precedence
-    ------------------------
     Each configuration parameter follows this resolution order:
         1. Explicit value passed to ArizeClient constructor (highest priority)
         2. Environment variable value
         3. Built-in default value (lowest priority)
 
-    Parameters
-    ----------
-    api_key : str
-        Arize API key for authentication. Required.
-        Environment variable: ARIZE_API_KEY
-        Default: None (must be provided via argument or environment variable)
-
-    api_host : str
-        API endpoint host.
-        Environment variable: ARIZE_API_HOST
-        Default: "api.arize.com"
-
-    api_scheme : str
-        API endpoint scheme (http/https).
-        Environment variable: ARIZE_API_SCHEME
-        Default: "https"
-
-    otlp_host : str
-        OTLP (OpenTelemetry Protocol) endpoint host.
-        Environment variable: ARIZE_OTLP_HOST
-        Default: "otlp.arize.com"
-
-    otlp_scheme : str
-        OTLP endpoint scheme (http/https).
-        Environment variable: ARIZE_OTLP_SCHEME
-        Default: "https"
-
-    flight_host : str
-        Apache Arrow Flight endpoint host.
-        Environment variable: ARIZE_FLIGHT_HOST
-        Default: "flight.arize.com"
-
-    flight_port : int
-        Apache Arrow Flight endpoint port (1-65535).
-        Environment variable: ARIZE_FLIGHT_PORT
-        Default: 443
-
-    flight_scheme : str
-        Apache Arrow Flight endpoint scheme.
-        Environment variable: ARIZE_FLIGHT_SCHEME
-        Default: "grpc+tls"
-
-    pyarrow_max_chunksize : int
-        Maximum chunk size for PyArrow operations (1 to MAX_CHUNKSIZE).
-        Environment variable: ARIZE_MAX_CHUNKSIZE
-        Default: 10_000
-
-    request_verify : bool
-        Whether to verify SSL certificates for HTTP requests.
-        Environment variable: ARIZE_REQUEST_VERIFY
-        Default: True
-
-    stream_max_workers : int
-        Maximum number of worker threads for streaming operations (minimum: 1).
-        Environment variable: ARIZE_STREAM_MAX_WORKERS
-        Default: 8
-
-    stream_max_queue_bound : int
-        Maximum queue size for streaming operations (minimum: 1).
-        Environment variable: ARIZE_STREAM_MAX_QUEUE_BOUND
-        Default: 5000
-
-    max_http_payload_size_mb : float
-        Maximum HTTP payload size in megabytes (minimum: 1).
-        Environment variable: ARIZE_MAX_HTTP_PAYLOAD_SIZE_MB
-        Default: 100
-
-    arize_directory : str
-        Directory for Arize SDK files (cache, logs, etc.).
-        Environment variable: ARIZE_DIRECTORY
-        Default: "~/.arize"
-
-    enable_caching : bool
-        Whether to enable local caching.
-        Environment variable: ARIZE_ENABLE_CACHING
-        Default: True
-
-    region : Region
-        Arize region (e.g., US_CENTRAL, EU_WEST). When specified, overrides
-        individual host/port settings.
-        Environment variable: ARIZE_REGION
-        Default: Region.UNSPECIFIED
-
-    single_host : str
-        Single host to use for all endpoints. Overrides individual host settings.
-        Environment variable: ARIZE_SINGLE_HOST
-        Default: "" (not set)
-
-    single_port : int
-        Single port to use for all endpoints. Overrides individual port settings (0-65535).
-        Environment variable: ARIZE_SINGLE_PORT
-        Default: 0 (not set)
-
-    See Also:
-    --------
-    ArizeClient : Main client class that uses this configuration
+    Args:
+        api_key: Arize API key for authentication. Required.
+            Environment variable: ARIZE_API_KEY.
+            Default: None (must be provided via argument or environment variable).
+        api_host: API endpoint host.
+            Environment variable: ARIZE_API_HOST.
+            Default: "api.arize.com".
+        api_scheme: API endpoint scheme (http/https).
+            Environment variable: ARIZE_API_SCHEME.
+            Default: "https".
+        otlp_host: OTLP (OpenTelemetry Protocol) endpoint host.
+            Environment variable: ARIZE_OTLP_HOST.
+            Default: "otlp.arize.com".
+        otlp_scheme: OTLP endpoint scheme (http/https).
+            Environment variable: ARIZE_OTLP_SCHEME.
+            Default: "https".
+        flight_host: Apache Arrow Flight endpoint host.
+            Environment variable: ARIZE_FLIGHT_HOST.
+            Default: "flight.arize.com".
+        flight_port: Apache Arrow Flight endpoint port (1-65535).
+            Environment variable: ARIZE_FLIGHT_PORT.
+            Default: 443.
+        flight_scheme: Apache Arrow Flight endpoint scheme.
+            Environment variable: ARIZE_FLIGHT_SCHEME.
+            Default: "grpc+tls".
+        pyarrow_max_chunksize: Maximum chunk size for PyArrow operations (1 to MAX_CHUNKSIZE).
+            Environment variable: ARIZE_MAX_CHUNKSIZE.
+            Default: 10_000.
+        request_verify: Whether to verify SSL certificates for HTTP requests.
+            Environment variable: ARIZE_REQUEST_VERIFY.
+            Default: True.
+        stream_max_workers: Maximum number of worker threads for streaming operations (minimum: 1).
+            Environment variable: ARIZE_STREAM_MAX_WORKERS.
+            Default: 8.
+        stream_max_queue_bound: Maximum queue size for streaming operations (minimum: 1).
+            Environment variable: ARIZE_STREAM_MAX_QUEUE_BOUND.
+            Default: 5000.
+        max_http_payload_size_mb: Maximum HTTP payload size in megabytes (minimum: 1).
+            Environment variable: ARIZE_MAX_HTTP_PAYLOAD_SIZE_MB.
+            Default: 100.
+        arize_directory: Directory for Arize SDK files (cache, logs, etc.).
+            Environment variable: ARIZE_DIRECTORY.
+            Default: "~/.arize".
+        enable_caching: Whether to enable local caching.
+            Environment variable: ARIZE_ENABLE_CACHING.
+            Default: True.
+        region: Arize region (e.g., US_CENTRAL, EU_WEST). When specified, overrides
+            individual host/port settings.
+            Environment variable: ARIZE_REGION.
+            Default: Region.UNSPECIFIED.
+        single_host: Single host to use for all endpoints. Overrides individual host settings.
+            Environment variable: ARIZE_SINGLE_HOST.
+            Default: "" (not set).
+        single_port: Single port to use for all endpoints. Overrides individual port settings (0-65535).
+            Environment variable: ARIZE_SINGLE_PORT.
+            Default: 0 (not set).
     """
 
     api_key: str = field(

arize/datasets/client.py

@@ -30,17 +30,22 @@ logger = logging.getLogger(__name__)
 
 
 class DatasetsClient:
-    """Client for managing datasets including creation, retrieval, and example management."""
+    """Client for managing datasets including creation, retrieval, and example management.
 
-    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
-        """Create a datasets sub-client.
+    This class is primarily intended for internal use within the SDK. Users are
+    highly encouraged to access resource-specific functionality via
+    :class:`arize.ArizeClient`.
 
-        The datasets client is a thin wrapper around the generated REST API client,
-        using the shared generated API client owned by `SDKConfiguration`.
+    The datasets client is a thin wrapper around the generated REST API client,
+    using the shared generated API client owned by
+    :class:`arize.config.SDKConfiguration`.
+    """
 
+    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
+        """
         Args:
             sdk_config: Resolved SDK configuration.
-        """
+        """  # noqa: D205, D212
         self._sdk_config = sdk_config
 
         # Import at runtime so it's still lazy and extras-gated by the parent

arize/embeddings/nlp_generators.py

@@ -49,10 +49,13 @@ class EmbeddingGeneratorForNLPSequenceClassification(NLPEmbeddingGenerator):
     ) -> pd.Series:
         """Obtain embedding vectors from your text data using pre-trained large language models.
 
-        :param text_col: a pandas Series containing the different pieces of text.
-        :param class_label_col: if this column is passed, the sentence "The classification label
-        is <class_label>" will be appended to the text in the `text_col`.
-        :return: a pandas Series containing the embedding vectors.
+        Args:
+            text_col: A pandas Series containing the different pieces of text.
+            class_label_col: If this column is passed, the sentence "The classification label
+                is <class_label>" will be appended to the text in the `text_col`.
+
+        Returns:
+            A pandas Series containing the embedding vectors.
         """
         if not isinstance(text_col, pd.Series):
             raise TypeError("text_col must be a pandas Series")
@@ -110,8 +113,11 @@ class EmbeddingGeneratorForNLPSummarization(NLPEmbeddingGenerator):
     ) -> pd.Series:
         """Obtain embedding vectors from your text data using pre-trained large language models.
 
-        :param text_col: a pandas Series containing the different pieces of text.
-        :return: a pandas Series containing the embedding vectors.
+        Args:
+            text_col: A pandas Series containing the different pieces of text.
+
+        Returns:
+            A pandas Series containing the embedding vectors.
         """
         if not isinstance(text_col, pd.Series):
             raise TypeError("text_col must be a pandas Series")

arize/embeddings/tabular_generators.py

@@ -11,7 +11,7 @@ from arize.embeddings.constants import (
     IMPORT_ERROR_MESSAGE,
 )
 from arize.embeddings.usecases import UseCases
-from arize.types import is_list_of
+from arize.utils.types import is_list_of
 
 try:
     from datasets import Dataset
@@ -79,16 +79,19 @@ class EmbeddingGeneratorForTabularFeatures(NLPEmbeddingGenerator):
         Prompts are generated from your `selected_columns` and passed to a pre-trained
         large language model for embedding vector computation.
 
-        :param df: pandas DataFrame containing the tabular data, not all columns will be
-        considered, see `selected_columns`.
-        :param selected_columns: columns to be considered to construct the prompt to be passed to
-        the LLM.
-        :param col_name_map: mapping between selected column names and a more verbose description of
-        the name. This helps the LLM understand the features better.
-        :param return_prompt_col: if set to True, an extra pandas Series will be returned
-        containing the constructed prompts. Defaults to False.
-        :return: a pandas Series containing the embedding vectors and, if `return_prompt_col` is
-        set to True, a pandas Series containing the prompts created from tabular features.
+        Args:
+            df: Pandas DataFrame containing the tabular data. Not all columns will be
+                considered, see `selected_columns`.
+            selected_columns: Columns to be considered to construct the prompt to be passed to
+                the LLM.
+            col_name_map: Mapping between selected column names and a more verbose description of
+                the name. This helps the LLM understand the features better.
+            return_prompt_col: If set to True, an extra pandas Series will be returned
+                containing the constructed prompts. Defaults to False.
+
+        Returns:
+            A pandas Series containing the embedding vectors and, if `return_prompt_col` is
+            set to True, a pandas Series containing the prompts created from tabular features.
         """
         if col_name_map is None:
             col_name_map = {}

arize/experiments/__init__.py

@@ -1 +1,13 @@
 """Experiment tracking and evaluation functionality for the Arize SDK."""
+
+from arize.experiments.evaluators.types import (
+    EvaluationResult,
+    EvaluationResultFieldNames,
+)
+from arize.experiments.types import ExperimentTaskResultFieldNames
+
+__all__ = [
+    "EvaluationResult",
+    "EvaluationResultFieldNames",
+    "ExperimentTaskResultFieldNames",
+]

arize/experiments/client.py

@@ -50,17 +50,22 @@ logger = logging.getLogger(__name__)
 
 
 class ExperimentsClient:
-    """Client for managing experiments including creation, execution, and result tracking."""
+    """Client for managing experiments including creation, execution, and result tracking.
 
-    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
-        """Create an experiments sub-client.
+    This class is primarily intended for internal use within the SDK. Users are
+    highly encouraged to access resource-specific functionality via
+    :class:`arize.ArizeClient`.
 
-        The experiments client is a thin wrapper around the generated REST API client,
-        using the shared generated API client owned by `SDKConfiguration`.
+    The experiments client is a thin wrapper around the generated REST API client,
+    using the shared generated API client owned by
+    :class:`arize.config.SDKConfiguration`.
+    """
 
+    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
+        """
         Args:
             sdk_config: Resolved SDK configuration.
-        """
+        """  # noqa: D205, D212
         self._sdk_config = sdk_config
         from arize._generated import api_client as gen
 

arize/{models → ml}/batch_validation/errors.py

@@ -16,12 +16,12 @@ from arize.constants.ml import (
     MAX_TAG_LENGTH,
 )
 from arize.logging import log_a_list
-from arize.types import Environments, ModelTypes
+from arize.ml.types import Environments, ModelTypes
 
 if TYPE_CHECKING:
     from collections.abc import Iterable
 
-    from arize.types import Metrics
+    from arize.ml.types import Metrics
 
 
 class ValidationError(Exception, ABC):

arize/{models → ml}/batch_validation/validator.py

@@ -40,8 +40,8 @@ from arize.constants.ml import (
     MODEL_MAPPING_CONFIG,
 )
 from arize.logging import get_truncation_warning_message
-from arize.models.batch_validation import errors as err
-from arize.types import (
+from arize.ml.batch_validation import errors as err
+from arize.ml.types import (
     CATEGORICAL_MODEL_TYPES,
     NUMERIC_MODEL_TYPES,
     BaseSchema,
@@ -53,9 +53,11 @@ from arize.types import (
     ModelTypes,
     PromptTemplateColumnNames,
     Schema,
+    segments_intersect,
+)
+from arize.utils.types import (
     is_dict_of,
     is_iterable_of,
-    segments_intersect,
 )
 
 logger = logging.getLogger(__name__)