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

arize/spans/client.py

@@ -26,7 +26,7 @@ from arize.exceptions.base import (
 from arize.exceptions.models import MissingProjectNameError
 from arize.exceptions.spaces import MissingSpaceIDError
 from arize.logging import CtxAdapter
-from arize.types import Environments, SimilaritySearchParams
+from arize.ml.types import Environments
 from arize.utils.arrow import post_arrow_table
 from arize.utils.dataframe import (
     remove_extraneous_columns,
@@ -44,14 +44,18 @@ logger = logging.getLogger(__name__)
 
 
 class SpansClient:
-    """Client for logging LLM tracing spans and evaluations to Arize."""
+    """Client for logging LLM tracing spans and evaluations to Arize.
 
-    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
-        """Initialize the spans client with SDK configuration.
+    This class is primarily intended for internal use within the SDK. Users are
+    highly encouraged to access resource-specific functionality via
+    :class:`arize.ArizeClient`.
+    """
 
-        Args:
-            sdk_config: SDK configuration containing API endpoints and credentials.
+    def __init__(self, *, sdk_config: SDKConfiguration) -> None:
         """
+        Args:
+            sdk_config: Resolved SDK configuration.
+        """  # noqa: D205, D212
         self._sdk_config = sdk_config
 
     def log(
@@ -72,19 +76,19 @@ class SpansClient:
         successful delivery of records.
 
         Args:
-            space_id (str): The space ID where the project resides.
-            project_name (str): A unique name to identify your project in the Arize platform.
-            dataframe (pd.DataFrame): The dataframe containing the LLM traces.
-            evals_dataframe (pd.DataFrame, optional): A dataframe containing LLM evaluations data.
+            space_id: The space ID where the project resides.
+            project_name: A unique name to identify your project in the Arize platform.
+            dataframe: The dataframe containing the LLM traces.
+            evals_dataframe: A dataframe containing LLM evaluations data.
                 The evaluations are joined to their corresponding spans via a left outer join, i.e.,
                 using only `context.span_id` from the spans dataframe. Defaults to None.
-            datetime_format (str): format for the timestamp captured in the LLM traces.
+            datetime_format: format for the timestamp captured in the LLM traces.
                 Defaults to "%Y-%m-%dT%H:%M:%S.%f+00:00".
-            validate (bool, optional): When set to True, validation is run before sending data.
+            validate: When set to True, validation is run before sending data.
                 Defaults to True.
-            timeout (float, optional): You can stop waiting for a response after a given number
+            timeout: You can stop waiting for a response after a given number
                 of seconds with the timeout parameter. Defaults to None.
-            tmp_dir (str, optional): Temporary directory/file to store the serialized data in binary
+            tmp_dir: Temporary directory/file to store the serialized data in binary
                 before sending to Arize.
 
         Returns:
@@ -274,15 +278,15 @@ class SpansClient:
         each evaluation to its respective span.
 
         Args:
-            space_id (str): The space ID where the project resides.
-            project_name (str): A unique name to identify your project in the Arize platform.
-            dataframe (pd.DataFrame): A dataframe containing LLM evaluations data.
-            validate (bool, optional): When set to True, validation is run before sending data.
+            space_id: The space ID where the project resides.
+            project_name: A unique name to identify your project in the Arize platform.
+            dataframe: A dataframe containing LLM evaluations data.
+            validate: When set to True, validation is run before sending data.
                 Defaults to True.
-            force_http (bool, optional): Force the use of HTTP for data upload. Defaults to False.
-            timeout (float, optional): You can stop waiting for a response after a given number
+            force_http: Force the use of HTTP for data upload. Defaults to False.
+            timeout: You can stop waiting for a response after a given number
                 of seconds with the timeout parameter. Defaults to None.
-            tmp_dir (str, optional): Temporary directory/file to store the serialized data in binary
+            tmp_dir: Temporary directory/file to store the serialized data in binary
                 before sending to Arize.
         """
         from arize.spans.columns import EVAL_COLUMN_PATTERN, SPAN_SPAN_ID_COL
@@ -447,10 +451,10 @@ class SpansClient:
         `annotation.notes` column can be included for free-form text notes.
 
         Args:
-            space_id (str): The space ID where the project resides.
-            project_name (str): A unique name to identify your project in the Arize platform.
-            dataframe (pd.DataFrame): A dataframe containing LLM annotation data.
-            validate (bool, optional): When set to True, validation is run before sending data.
+            space_id: The space ID where the project resides.
+            project_name: A unique name to identify your project in the Arize platform.
+            dataframe: A dataframe containing LLM annotation data.
+            validate: When set to True, validation is run before sending data.
                 Defaults to True.
         """
         from arize.spans.columns import (
@@ -661,6 +665,7 @@ class SpansClient:
         This method is only supported for LLM model types.
 
         The dataframe must contain a column `context.span_id` to identify spans and either:
+
         1. A column with JSON patch documents (specified by patch_document_column_name), or
         2. One or more columns with prefix `attributes.metadata.` that will be automatically
            converted to a patch document (e.g., `attributes.metadata.tag` → `{"tag": value}`).
@@ -668,7 +673,8 @@ class SpansClient:
         If both methods are used, the explicit patch document is applied after the individual field updates.
         The patches will be applied to the `attributes.metadata` field of each span.
 
-        **Type Handling:**
+        Type Handling:
+
         - The client primarily supports string, integer, and float data types.
         - Boolean values are converted to string representations.
         - Nested JSON objects and arrays are serialized to JSON strings during transmission.
@@ -685,12 +691,14 @@ class SpansClient:
 
         Returns:
             Dictionary containing update results with the following keys:
+
                 - spans_processed: Total number of spans in the input dataframe
                 - spans_updated: Count of successfully updated span metadata records
                 - spans_failed: Count of spans that failed to update
                 - errors: List of dictionaries with 'span_id' and 'error_message' keys for each failed span
 
-            Error types from the server include:
+                Error types from the server include:
+
                 - parse_failure: Failed to parse JSON metadata
                 - patch_failure: Failed to apply JSON patch
                 - type_conflict: Type conflict in metadata
@@ -699,58 +707,60 @@ class SpansClient:
                 - druid_rejection: Backend rejected the update
 
         Raises:
-            AuthError: When API key or space ID is missing
-            ValidationFailure: When validation of the dataframe or values fails
-            ImportError: When required tracing dependencies are missing
-            ArrowInvalid: When the dataframe cannot be converted to Arrow format
-            RuntimeError: If the request fails or no response is received
-
-        Example:
-            ```python
-            # Method 1: Using a patch document
-            df = pd.DataFrame(
-                {
-                    "context.span_id": ["span1", "span2"],
-                    "patch_document": [
-                        {"tag": "important"},
-                        {"priority": "high"},
-                    ],
-                }
-            )
-
-            # Method 2: Using direct field columns
-            df = pd.DataFrame(
-                {
-                    "context.span_id": ["span1", "span2"],
-                    "attributes.metadata.tag": ["important", "standard"],
-                    "attributes.metadata.priority": ["high", "medium"],
-                }
-            )
-
-            # Method 3: Combining both approaches
-            df = pd.DataFrame(
-                {
-                    "context.span_id": ["span1"],
-                    "attributes.metadata.tag": ["important"],
-                    "patch_document": [
-                        {"priority": "high"}
-                    ],  # This will override any conflicting fields
-                }
-            )
-
-            # Method 4: Setting fields to null
-            df = pd.DataFrame(
-                {
-                    "context.span_id": ["span1"],
-                    "attributes.metadata.old_field": [
-                        None
-                    ],  # Sets field to JSON null
-                    "patch_document": [
-                        {"other_field": None}
-                    ],  # Also sets field to JSON null
-                }
-            )
-            ```
+            AuthError: When API key or space ID is missing.
+            ValidationFailure: When validation of the dataframe or values fails.
+            ImportError: When required tracing dependencies are missing.
+            ArrowInvalid: When the dataframe cannot be converted to Arrow format.
+            RuntimeError: If the request fails or no response is received.
+
+        Examples:
+            Method 1: Using a patch document
+
+            >>> df = pd.DataFrame(
+            ...     {
+            ...         "context.span_id": ["span1", "span2"],
+            ...         "patch_document": [
+            ...             {"tag": "important"},
+            ...             {"priority": "high"},
+            ...         ],
+            ...     }
+            ... )
+
+            Method 2: Using direct field columns
+
+            >>> df = pd.DataFrame(
+            ...     {
+            ...         "context.span_id": ["span1", "span2"],
+            ...         "attributes.metadata.tag": ["important", "standard"],
+            ...         "attributes.metadata.priority": ["high", "medium"],
+            ...     }
+            ... )
+
+            Method 3: Combining both approaches
+
+            >>> df = pd.DataFrame(
+            ...     {
+            ...         "context.span_id": ["span1"],
+            ...         "attributes.metadata.tag": ["important"],
+            ...         "patch_document": [
+            ...             {"priority": "high"}
+            ...         ],  # Overrides conflicting fields
+            ...     }
+            ... )
+
+            Method 4: Setting fields to null
+
+            >>> df = pd.DataFrame(
+            ...     {
+            ...         "context.span_id": ["span1"],
+            ...         "attributes.metadata.old_field": [
+            ...             None
+            ...         ],  # Sets field to JSON null
+            ...         "patch_document": [
+            ...             {"other_field": None}
+            ...         ],  # Also sets field to JSON null
+            ...     }
+            ... )
         """
         # Import validation modules
         from arize.spans.columns import SPAN_SPAN_ID_COL
@@ -992,7 +1002,6 @@ class SpansClient:
         end_time: datetime,
         where: str = "",
         columns: list | None = None,
-        similarity_search_params: SimilaritySearchParams | None = None,
         stream_chunk_size: int | None = None,
     ) -> pd.DataFrame:
         """Export span data from Arize to a pandas DataFrame.
@@ -1002,7 +1011,6 @@ class SpansClient:
         WHERE clauses and similarity search for semantic retrieval.
 
         Returns:
-        -------
             pd.DataFrame: DataFrame containing the requested span data with columns
                 for span metadata, attributes, events, and any custom fields.
         """
@@ -1025,7 +1033,6 @@ class SpansClient:
                 end_time=end_time,
                 where=where,
                 columns=columns,
-                similarity_search_params=similarity_search_params,
                 stream_chunk_size=stream_chunk_size,
             )
 
@@ -1039,7 +1046,6 @@ class SpansClient:
         end_time: datetime,
         where: str = "",
         columns: list | None = None,
-        similarity_search_params: SimilaritySearchParams | None = None,
         stream_chunk_size: int | None = None,
     ) -> None:
         """Export span data from Arize to a Parquet file.
@@ -1069,7 +1075,6 @@ class SpansClient:
                 end_time=end_time,
                 where=where,
                 columns=columns,
-                similarity_search_params=similarity_search_params,
                 stream_chunk_size=stream_chunk_size,
             )
 

arize/spans/conversion.py

@@ -35,10 +35,17 @@ def convert_timestamps(df: pd.DataFrame, fmt: str = "") -> pd.DataFrame:
 
 def _datetime_to_ns(dt: object, fmt: str) -> int:
     if isinstance(dt, str):
-        return int(
-            datetime.strptime(dt, fmt).replace(tzinfo=timezone.utc).timestamp()
-            * 1e9
-        )
+        # Try ISO 8601 with timezone first
+        try:
+            parsed = datetime.fromisoformat(dt)
+            if parsed.tzinfo is None:
+                # If no timezone, assume UTC
+                parsed = parsed.replace(tzinfo=timezone.utc)
+        except ValueError:
+            # Fall back to custom format
+            parsed = datetime.strptime(dt, fmt).replace(tzinfo=timezone.utc)
+
+        return int(parsed.timestamp() * 1e9)
     if isinstance(dt, datetime):
         return int(datetime.timestamp(dt) * 1e9)
     if isinstance(dt, pd.Timestamp):

arize/spans/validation/common/value_validation.py

@@ -25,7 +25,7 @@ from arize.spans.validation.common.errors import (
     InvalidStringValueNotAllowedInColumn,
     InvalidTimestampValueInColumn,
 )
-from arize.types import is_json_str
+from arize.utils.types import is_json_str
 
 logger = logging.getLogger(__name__)
 

arize/spans/validation/spans/dataframe_form_validation.py

@@ -12,7 +12,7 @@ from arize.spans.conversion import is_missing_value
 from arize.spans.validation.common.errors import (
     InvalidDataFrameColumnContentTypes,
 )
-from arize.types import is_array_of, is_dict_of, is_list_of
+from arize.utils.types import is_array_of, is_dict_of, is_list_of
 
 logger = logging.getLogger(__name__)
 

arize/spans/validation/spans/value_validation.py

@@ -7,6 +7,7 @@ from typing import TYPE_CHECKING
 
 from arize.constants import spans as tracing_constants
 from arize.constants.ml import MAX_EMBEDDING_DIMENSIONALITY
+from arize.ml.types import StatusCodes
 from arize.spans import columns as tracing_cols
 from arize.spans.validation.common import value_validation
 from arize.spans.validation.common.errors import (
@@ -15,7 +16,7 @@ from arize.spans.validation.common.errors import (
     InvalidEventValueInColumn,
     InvalidLLMMessageValueInColumn,
 )
-from arize.types import StatusCodes, is_dict_of, is_json_str
+from arize.utils.types import is_dict_of, is_json_str
 
 if TYPE_CHECKING:
     import pandas as pd

arize/utils/dataframe.py

@@ -4,7 +4,7 @@ import re
 
 import pandas as pd
 
-from arize.types import BaseSchema
+from arize.ml.types import BaseSchema
 
 
 # Resets the dataframe index if it is not a RangeIndex

arize/utils/online_tasks/dataframe_preprocessor.py

@@ -122,7 +122,7 @@ def extract_nested_data_to_column(
 def _introspect_arize_attribute(value: object, attribute: str) -> object:
     """Recursively drill into `value` following the dot-delimited `attribute`.
 
-    Example:
+    Examples:
         value: [{'message.role': 'assistant', 'message.content': 'The capital of China is Beijing.'}]
         attribute: "0.message.content"
         Returns: 'The capital of China is Beijing.'
@@ -132,7 +132,6 @@ def _introspect_arize_attribute(value: object, attribute: str) -> object:
       - Parses JSON strings
       - Converts NumPy arrays to lists
       - Allows dotted keys (e.g. "message.content") by combining parts
-
     """
     if not attribute:
         return value
@@ -195,10 +194,10 @@ def _parse_value(
     idx = _try_int(key)
     if idx is not None:
         # Must be a tuple or list (_ensure_deserialized() already casts numpy arrays to python lists)
-        if isinstance(current_value, (list, tuple)):
-            if 0 <= idx < len(current_value):
-                return (current_value[idx], num_parts_processed)
-            return (None, num_parts_processed)
+        if isinstance(current_value, list | tuple) and 0 <= idx < len(
+            current_value
+        ):
+            return (current_value[idx], num_parts_processed)
         return (None, num_parts_processed)
 
     # 2) Try dict approach

arize/utils/types.py

@@ -0,0 +1,105 @@
+"""Common type definitions and data models used across the Arize SDK."""
+
+import json
+from collections.abc import Iterable, Sequence
+from typing import (
+    TypeVar,
+)
+
+import numpy as np
+
+
+def is_json_str(s: str) -> bool:
+    """Check if a string is valid JSON.
+
+    Args:
+        s: The string to validate.
+
+    Returns:
+        True if the string is valid JSON, False otherwise.
+    """
+    try:
+        json.loads(s)
+    except ValueError:
+        return False
+    except TypeError:
+        return False
+    return True
+
+
+T = TypeVar("T", bound=type)
+
+
+def is_array_of(arr: Sequence[object], tp: T) -> bool:
+    """Check if a value is a numpy array with all elements of a specific type.
+
+    Args:
+        arr: The sequence to check.
+        tp: The expected type for all elements.
+
+    Returns:
+        True if arr is a numpy array and all elements are of type tp.
+    """
+    return isinstance(arr, np.ndarray) and all(isinstance(x, tp) for x in arr)
+
+
+def is_list_of(lst: Sequence[object], tp: T) -> bool:
+    """Check if a value is a list with all elements of a specific type.
+
+    Args:
+        lst: The sequence to check.
+        tp: The expected type for all elements.
+
+    Returns:
+        True if lst is a list and all elements are of type tp.
+    """
+    return isinstance(lst, list) and all(isinstance(x, tp) for x in lst)
+
+
+def is_iterable_of(lst: Sequence[object], tp: T) -> bool:
+    """Check if a value is an iterable with all elements of a specific type.
+
+    Args:
+        lst: The sequence to check.
+        tp: The expected type for all elements.
+
+    Returns:
+        True if lst is an iterable and all elements are of type tp.
+    """
+    return isinstance(lst, Iterable) and all(isinstance(x, tp) for x in lst)
+
+
+def is_dict_of(
+    d: dict[object, object],
+    key_allowed_types: T,
+    value_allowed_types: T = (),
+    value_list_allowed_types: T = (),
+) -> bool:
+    """Method to check types are valid for dictionary.
+
+    Args:
+        d: Dictionary itself.
+        key_allowed_types: All allowed types for keys of dictionary.
+        value_allowed_types: All allowed types for values of dictionary.
+        value_list_allowed_types: If value is a list, these are the allowed
+            types for value list.
+
+    Returns:
+        True if the data types of dictionary match the types specified by the
+            arguments, false otherwise.
+    """
+    if value_list_allowed_types and not isinstance(
+        value_list_allowed_types, tuple
+    ):
+        value_list_allowed_types = (value_list_allowed_types,)
+
+    return (
+        isinstance(d, dict)
+        and all(isinstance(k, key_allowed_types) for k in d)
+        and all(
+            isinstance(v, value_allowed_types)
+            or any(is_list_of(v, t) for t in value_list_allowed_types)
+            for v in d.values()
+            if value_allowed_types or value_list_allowed_types
+        )
+    )

arize/version.py

@@ -1,3 +1,3 @@
 """Version information for the Arize SDK."""
 
-__version__ = "8.0.0a23"
+__version__ = "8.0.0b0"

{arize-8.0.0a23.dist-info → arize-8.0.0b0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: arize
-Version: 8.0.0a23
+Version: 8.0.0b0
 Summary: A helper library to interact with Arize AI APIs
 Project-URL: Homepage, https://arize.com
 Project-URL: Documentation, https://docs.arize.com/arize
@@ -37,6 +37,12 @@ Requires-Dist: torch<3,>=1.13; extra == 'auto-embeddings'
 Requires-Dist: transformers<5,>=4.25; extra == 'auto-embeddings'
 Provides-Extra: datasets-experiments
 Requires-Dist: numpy>=2.0.0; extra == 'datasets-experiments'
+Requires-Dist: openinference-semantic-conventions<1,>=0.1.21; extra == 'datasets-experiments'
+Requires-Dist: opentelemetry-api>=1.38.0; extra == 'datasets-experiments'
+Requires-Dist: opentelemetry-exporter-otlp-proto-common>=1.38.0; extra == 'datasets-experiments'
+Requires-Dist: opentelemetry-exporter-otlp-proto-grpc>=1.38.0; extra == 'datasets-experiments'
+Requires-Dist: opentelemetry-proto>=1.38.0; extra == 'datasets-experiments'
+Requires-Dist: opentelemetry-sdk>=1.38.0; extra == 'datasets-experiments'
 Requires-Dist: pydantic; extra == 'datasets-experiments'
 Requires-Dist: wrapt<2.0.0,>=1.0.0; extra == 'datasets-experiments'
 Provides-Extra: dev
@@ -261,7 +267,7 @@ Use `arize.models` to interact with ML models: log ML data (traininv, validation
 
 ```python
 from arize import ArizeClient
-from arize.types import ModelTypes, Environments
+from arize.ml.types import ModelTypes, Environments
 
 client = ArizeClient(api_key=API_KEY)
 SPACE_ID = "<your-space-id>"
@@ -286,14 +292,14 @@ response = client.models.log_stream(
 
 ```python
 from arize import ArizeClient
-from arize.types import ModelTypes, Environments
+from arize.ml.types import ModelTypes, Environments
 
 client = ArizeClient(api_key=API_KEY)
 SPACE_ID = "<your-space-id>"
 MODEL_NAME = "<your-model-name>"
 MODEL_VERSION = "1.0"
 
-from arize.types import Schema, EmbeddingColumnNames, ObjectDetectionColumnNames, ModelTypes, Environments
+from arize.ml.types import Schema, EmbeddingColumnNames, ObjectDetectionColumnNames, ModelTypes, Environments
 
 tags = ["drift_type"]
 embedding_feature_column_names = {