arize-phoenix 11.6.2__tar.gz → 11.8.0__tar.gz

{arize_phoenix-11.6.2 → arize_phoenix-11.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: arize-phoenix
-Version: 11.6.2
+Version: 11.8.0
 Summary: AI Observability and Evaluation
 Project-URL: Documentation, https://arize.com/docs/phoenix/
 Project-URL: Issues, https://github.com/Arize-ai/phoenix/issues
@@ -40,7 +40,7 @@ Requires-Dist: opentelemetry-sdk
 Requires-Dist: opentelemetry-semantic-conventions
 Requires-Dist: pandas>=1.0
 Requires-Dist: prometheus-client
-Requires-Dist: protobuf<6.0,>=4.25.8
+Requires-Dist: protobuf>=4.25.8
 Requires-Dist: psutil
 Requires-Dist: pyarrow
 Requires-Dist: pydantic>=2.1.0
@@ -65,13 +65,13 @@ Requires-Dist: fast-hdbscan>=0.2.0; extra == 'container'
 Requires-Dist: google-generativeai; extra == 'container'
 Requires-Dist: numba>=0.60.0; extra == 'container'
 Requires-Dist: openai>=1.0.0; extra == 'container'
-Requires-Dist: opentelemetry-exporter-otlp; extra == 'container'
-Requires-Dist: opentelemetry-instrumentation-fastapi; extra == 'container'
-Requires-Dist: opentelemetry-instrumentation-grpc; extra == 'container'
-Requires-Dist: opentelemetry-instrumentation-sqlalchemy; extra == 'container'
-Requires-Dist: opentelemetry-proto>=1.12.0; extra == 'container'
-Requires-Dist: opentelemetry-sdk; extra == 'container'
-Requires-Dist: opentelemetry-semantic-conventions; extra == 'container'
+Requires-Dist: opentelemetry-exporter-otlp==1.33.1; extra == 'container'
+Requires-Dist: opentelemetry-instrumentation-fastapi==0.54b1; extra == 'container'
+Requires-Dist: opentelemetry-instrumentation-grpc==0.54b1; extra == 'container'
+Requires-Dist: opentelemetry-instrumentation-sqlalchemy==0.54b1; extra == 'container'
+Requires-Dist: opentelemetry-proto==1.33.1; extra == 'container'
+Requires-Dist: opentelemetry-sdk==1.33.1; extra == 'container'
+Requires-Dist: opentelemetry-semantic-conventions==0.54b1; extra == 'container'
 Requires-Dist: prometheus-client; extra == 'container'
 Requires-Dist: py-grpc-prometheus; extra == 'container'
 Requires-Dist: strawberry-graphql[opentelemetry]==0.270.1; extra == 'container'
@@ -205,6 +205,7 @@ The `arize-phoenix` package includes the entire Phoenix platfom. However if you
 | [arize-phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-client)       | Python [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-client)](https://pypi.org/project/arize-phoenix-client/)                 | Lightweight client for interacting with the Phoenix server via its OpenAPI REST interface         |
 | [arize-phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-evals)         | Python [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-evals)](https://pypi.org/project/arize-phoenix-evals/)                   | Tooling to evaluate LLM applications including RAG relevance, answer relevance, and more          |
 | [@arizeai/phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-client) | JavaScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-client)](https://www.npmjs.com/package/@arizeai/phoenix-client) | Client for the Arize Phoenix API                                                                  |
+| [@arizeai/phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-evals)   | TypeScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-evals)](https://www.npmjs.com/package/@arizeai/phoenix-evals)   | TypeScript evaluation library for LLM applications (alpha release)                                |
 | [@arizeai/phoenix-mcp](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-mcp)       | JavaScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-mcp)](https://www.npmjs.com/package/@arizeai/phoenix-mcp)       | MCP server implementation for Arize Phoenix providing unified interface to Phoenix's capabilities |
 
 ## Tracing Integrations
@@ -248,9 +249,20 @@ Phoenix is built on top of OpenTelemetry and is vendor, language, and framework
 | [BeeAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/beeai)                 | `@arizeai/openinference-instrumentation-beeai`     | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-vercel)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-beeai)                            |
 | [Mastra](https://arize.com/docs/phoenix/integrations/mastra)                               | `@arizeai/openinference-mastra`                    | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-mastra.svg)](https://www.npmjs.com/package/@arizeai/openinference-mastra)                                       |
 
+### Java Integrations
+
+| Integration                                                                                                                       | Package                                     | Version Badge                                                                                                                                                                                                 |
+| --------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| [LangChain4j](https://github.com/Arize-ai/openinference/tree/main/java/instrumentation/openinference-instrumentation-langchain4j) | `openinference-instrumentation-langchain4j` | [![Maven Central](https://img.shields.io/maven-central/v/com.arize/openinference-instrumentation-langchain4j.svg)](https://central.sonatype.com/artifact/com.arize/openinference-instrumentation-langchain4j) |
+
 ### Platforms
 
-Phoenix has native integrations with [LangFlow](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow), LiteLLM Proxy, and [BeeAI](https://docs.beeai.dev/observability/agents-traceability).
+| Platform                                                                                                 | Description                                                    | Docs                                                                                                          |
+| -------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------- |
+| [Dify](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix)         | Open-source LLM app development platform                       | [Integration Guide](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix) |
+| [LangFlow](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow)                         | Visual framework for building multi-agent and RAG applications | [Integration Guide](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow)                     |
+| [BeeAI](https://docs.beeai.dev/observability/agents-traceability)                                        | AI agent framework with built-in observability                 | [Integration Guide](https://docs.beeai.dev/observability/agents-traceability)                                 |
+| [LiteLLM Proxy](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy) | Proxy server for LLMs                                          | [Integration Guide](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy)  |
 
 ## Community
 

{arize_phoenix-11.6.2 → arize_phoenix-11.8.0}/README.md

@@ -72,6 +72,7 @@ The `arize-phoenix` package includes the entire Phoenix platfom. However if you
 | [arize-phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-client)       | Python [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-client)](https://pypi.org/project/arize-phoenix-client/)                 | Lightweight client for interacting with the Phoenix server via its OpenAPI REST interface         |
 | [arize-phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-evals)         | Python [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-evals)](https://pypi.org/project/arize-phoenix-evals/)                   | Tooling to evaluate LLM applications including RAG relevance, answer relevance, and more          |
 | [@arizeai/phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-client) | JavaScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-client)](https://www.npmjs.com/package/@arizeai/phoenix-client) | Client for the Arize Phoenix API                                                                  |
+| [@arizeai/phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-evals)   | TypeScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-evals)](https://www.npmjs.com/package/@arizeai/phoenix-evals)   | TypeScript evaluation library for LLM applications (alpha release)                                |
 | [@arizeai/phoenix-mcp](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-mcp)       | JavaScript [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-mcp)](https://www.npmjs.com/package/@arizeai/phoenix-mcp)       | MCP server implementation for Arize Phoenix providing unified interface to Phoenix's capabilities |
 
 ## Tracing Integrations
@@ -115,9 +116,20 @@ Phoenix is built on top of OpenTelemetry and is vendor, language, and framework
 | [BeeAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/beeai)                 | `@arizeai/openinference-instrumentation-beeai`     | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-vercel)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-beeai)                            |
 | [Mastra](https://arize.com/docs/phoenix/integrations/mastra)                               | `@arizeai/openinference-mastra`                    | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-mastra.svg)](https://www.npmjs.com/package/@arizeai/openinference-mastra)                                       |
 
+### Java Integrations
+
+| Integration                                                                                                                       | Package                                     | Version Badge                                                                                                                                                                                                 |
+| --------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| [LangChain4j](https://github.com/Arize-ai/openinference/tree/main/java/instrumentation/openinference-instrumentation-langchain4j) | `openinference-instrumentation-langchain4j` | [![Maven Central](https://img.shields.io/maven-central/v/com.arize/openinference-instrumentation-langchain4j.svg)](https://central.sonatype.com/artifact/com.arize/openinference-instrumentation-langchain4j) |
+
 ### Platforms
 
-Phoenix has native integrations with [LangFlow](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow), LiteLLM Proxy, and [BeeAI](https://docs.beeai.dev/observability/agents-traceability).
+| Platform                                                                                                 | Description                                                    | Docs                                                                                                          |
+| -------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------- |
+| [Dify](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix)         | Open-source LLM app development platform                       | [Integration Guide](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix) |
+| [LangFlow](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow)                         | Visual framework for building multi-agent and RAG applications | [Integration Guide](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow)                     |
+| [BeeAI](https://docs.beeai.dev/observability/agents-traceability)                                        | AI agent framework with built-in observability                 | [Integration Guide](https://docs.beeai.dev/observability/agents-traceability)                                 |
+| [LiteLLM Proxy](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy) | Proxy server for LLMs                                          | [Integration Guide](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy)  |
 
 ## Community
 

{arize_phoenix-11.6.2 → arize_phoenix-11.8.0}/pyproject.toml

@@ -34,7 +34,7 @@ dependencies = [
   "typing-extensions>=4.6",
   "scipy",
   "wrapt>=1.17.2",
-  "protobuf>=4.25.8, <6.0",
+  "protobuf>=4.25.8",
   "grpcio",
   "grpc-interceptor",
   "tqdm",
@@ -131,13 +131,15 @@ container = [
   "openai>=1.0.0",
   "azure-identity",
   "aiohttp",  # used by azure-identity via azure.core.pipeline.transport
-  "opentelemetry-sdk",
-  "opentelemetry-proto>=1.12.0",
-  "opentelemetry-exporter-otlp",
-  "opentelemetry-semantic-conventions",
-  "opentelemetry-instrumentation-fastapi",
-  "opentelemetry-instrumentation-sqlalchemy",
-  "opentelemetry-instrumentation-grpc",
+  # Note: when updating opentelemetry dependencies, all packages must use versions
+  # released on the same date (e.g., 1.33.1/0.54b1) to ensure compatibility.
+  "opentelemetry-sdk==1.33.1",
+  "opentelemetry-proto==1.33.1",
+  "opentelemetry-exporter-otlp==1.33.1",
+  "opentelemetry-semantic-conventions==0.54b1",
+  "opentelemetry-instrumentation-fastapi==0.54b1",  # https://github.com/open-telemetry/opentelemetry-python-contrib/issues/3637
+  "opentelemetry-instrumentation-sqlalchemy==0.54b1",
+  "opentelemetry-instrumentation-grpc==0.54b1",
   "py-grpc-prometheus",
   "strawberry-graphql[opentelemetry]==0.270.1",
   "uvloop; platform_system != 'Windows'",

{arize_phoenix-11.6.2 → arize_phoenix-11.8.0}/src/phoenix/config.py

@@ -246,7 +246,14 @@ The URL to use for redirecting to a management interface that may be hosting Pho
 the current user is within PHOENIX_ADMINS, a link will be added to the navigation menu to return to
 this URL.
 """
+ENV_PHOENIX_SUPPORT_EMAIL = "PHOENIX_SUPPORT_EMAIL"
+"""
+The support email address to display in error messages and notifications.
 
+When set, this email will be included in error messages for insufficient storage
+conditions and database usage notification emails, providing users with a direct
+contact for assistance. If not set, error messages will not include contact information.
+"""
 
 # SMTP settings
 ENV_PHOENIX_SMTP_HOSTNAME = "PHOENIX_SMTP_HOSTNAME"
@@ -1599,6 +1606,31 @@ def get_env_management_url() -> Optional[str]:
     return getenv(ENV_PHOENIX_MANAGEMENT_URL)
 
 
+def get_env_support_email() -> Optional[str]:
+    """
+    Get the support email address from the PHOENIX_SUPPORT_EMAIL environment variable.
+
+    Returns:
+        The support email address if set, None otherwise.
+    """
+    return getenv(ENV_PHOENIX_SUPPORT_EMAIL)
+
+
+def validate_env_support_email() -> None:
+    """
+    Validate the support email address configured in PHOENIX_SUPPORT_EMAIL.
+
+    Raises:
+        ValueError: If the email address is invalid.
+    """
+    if not (email := get_env_support_email()):
+        return
+    try:
+        validate_email(email, check_deliverability=False)
+    except EmailNotValidError as e:
+        raise ValueError(f"Invalid email in {ENV_PHOENIX_SUPPORT_EMAIL}: '{email}'") from e
+
+
 def verify_server_environment_variables() -> None:
     """Verify that the environment variables are set correctly. Raises an error otherwise."""
     get_env_root_url()
@@ -1607,6 +1639,7 @@ def verify_server_environment_variables() -> None:
     get_env_database_allocated_storage_capacity_gibibytes()
     get_env_database_usage_email_warning_threshold_percentage()
     get_env_database_usage_insertion_blocking_threshold_percentage()
+    validate_env_support_email()
 
     # Notify users about deprecated environment variables if they are being used.
     if os.getenv("PHOENIX_ENABLE_WEBSOCKETS") is not None:

arize_phoenix-11.8.0/src/phoenix/datetime_utils.py

@@ -0,0 +1,226 @@
+from datetime import datetime, timedelta, timezone, tzinfo
+from typing import Any, Iterator, Literal, Optional, cast
+
+import pandas as pd
+import pytz
+from pandas import Timestamp, to_datetime
+from pandas.core.dtypes.common import (
+    is_datetime64_any_dtype,
+    is_datetime64tz_dtype,
+    is_numeric_dtype,
+    is_object_dtype,
+)
+from typing_extensions import assert_never
+
+_LOCAL_TIMEZONE = datetime.now(timezone.utc).astimezone().tzinfo
+
+
+def local_now() -> datetime:
+    return datetime.now(timezone.utc).astimezone(tz=_LOCAL_TIMEZONE)
+
+
+def normalize_datetime(
+    dt: Optional[datetime],
+    tz: Optional[tzinfo] = None,
+) -> Optional[datetime]:
+    """
+    If the input datetime is timezone-naive, it is localized as local timezone
+    unless tzinfo is specified.
+    """
+    if not isinstance(dt, datetime):
+        return None
+    if not is_timezone_aware(dt):
+        dt = dt.replace(tzinfo=tz if tz else _LOCAL_TIMEZONE)
+    return dt.astimezone(timezone.utc)
+
+
+def normalize_timestamps(
+    timestamps: "pd.Series[Any]",
+) -> "pd.Series[Timestamp]":
+    """
+    If the input timestamps contains a Unix or datetime timestamp or ISO8601
+    timestamp strings column, it is converted to UTC timezone-aware timestamp.
+    If a timestamp is timezone-naive, it is localized as per local timezone
+    and then converted to UTC.
+    """
+    if is_numeric_dtype(timestamps):
+        return to_datetime(timestamps, unit="s", utc=True)
+    if is_datetime64tz_dtype(timestamps):
+        return timestamps.dt.tz_convert(pytz.utc)
+    if is_datetime64_any_dtype(timestamps):
+        return timestamps.dt.tz_localize(
+            datetime.now().astimezone().tzinfo,
+        ).dt.tz_convert(
+            timezone.utc,
+        )
+    if is_object_dtype(timestamps):
+        timestamps = to_datetime(timestamps)
+        if timestamps.dt.tz is None:
+            timestamps = timestamps.dt.tz_localize(
+                datetime.now().astimezone().tzinfo,
+            )
+        return timestamps.dt.tz_convert(
+            timezone.utc,
+        )
+    raise ValueError(
+        "When provided, input timestamp column must have numeric or datetime dtype, "
+        f"but found {timestamps.dtype} instead."
+    )
+
+
+def floor_to_minute(dt: datetime) -> datetime:
+    """Floor datetime to the minute by taking a round-trip through string
+    format because there isn't always an available function to strip the
+    nanoseconds if present."""
+    try:
+        dt_as_string = dt.astimezone(
+            timezone.utc,
+        ).strftime(
+            MINUTE_DATETIME_FORMAT,
+        )
+    except ValueError:
+        # NOTE: as of Python 3.8.16, pandas 1.5.3:
+        # >>> isinstance(pd.NaT, datetime.datetime)
+        # True
+        return cast(datetime, pd.NaT)
+    return datetime.strptime(
+        dt_as_string,
+        MINUTE_DATETIME_FORMAT,
+    ).astimezone(
+        timezone.utc,
+    )
+
+
+MINUTE_DATETIME_FORMAT = "%Y-%m-%dT%H:%M:00%z"
+
+
+def right_open_time_range(
+    min_time: Optional[datetime],
+    max_time: Optional[datetime],
+) -> tuple[Optional[datetime], Optional[datetime]]:
+    """
+    First adds one minute to `max_time`, because time intervals are right
+    open and one minute is the smallest interval allowed, then rounds down
+    the times to the nearest minute.
+    """
+    return (
+        floor_to_minute(min_time) if min_time else None,
+        floor_to_minute(max_time + timedelta(minutes=1)) if max_time else None,
+    )
+
+
+def is_timezone_aware(dt: datetime) -> bool:
+    """
+    Returns True if the datetime is timezone-aware, False otherwise.
+    """
+    return dt.tzinfo is not None and dt.tzinfo.utcoffset(dt) is not None
+
+
+def get_timestamp_range(
+    start_time: datetime,
+    end_time: datetime,
+    stride: Literal["minute", "hour", "day", "week", "month", "year"] = "minute",
+    utc_offset_minutes: int = 0,
+) -> Iterator[datetime]:
+    """
+    Generate a sequence of datetime objects at regular intervals between start and end times.
+
+    This function creates time intervals by rounding down the start time to the nearest
+    stride boundary in the specified timezone, then yielding timestamps at regular
+    intervals until reaching the end time. All returned timestamps are in UTC.
+
+    Args:
+        start_time: The starting datetime (inclusive after rounding down to stride boundary).
+                   Must be timezone-aware.
+        end_time: The ending datetime (exclusive). Must be timezone-aware.
+        stride: The interval between generated timestamps. Options:
+               - "minute": Generate timestamps every minute
+               - "hour": Generate timestamps every hour
+               - "day": Generate timestamps every day at midnight
+               - "week": Generate timestamps every week at Monday midnight
+               - "month": Generate timestamps on the 1st of each month at midnight
+               - "year": Generate timestamps on January 1st of each year at midnight
+        utc_offset_minutes: Timezone offset in minutes from UTC. Used to determine
+                           the correct stride boundaries in local time. Positive values
+                           are east of UTC, negative values are west of UTC.
+
+    Returns:
+        Iterator of datetime objects in UTC timezone, spaced at the specified stride
+        interval. The first timestamp is rounded down to the nearest stride boundary
+        in the local timezone (considering utc_offset_minutes).
+
+    Examples:
+        >>> from datetime import datetime, timezone
+        >>> start = datetime(2024, 1, 1, 12, 30, 45, tzinfo=timezone.utc)
+        >>> end = datetime(2024, 1, 1, 12, 33, 0, tzinfo=timezone.utc)
+        >>> list(get_timestamp_range(start, end, "minute"))
+        [datetime(2024, 1, 1, 12, 30, tzinfo=timezone.utc),
+         datetime(2024, 1, 1, 12, 31, tzinfo=timezone.utc),
+         datetime(2024, 1, 1, 12, 32, tzinfo=timezone.utc)]
+
+        >>> # Week stride rounds down to Monday
+        >>> start = datetime(2024, 1, 10, 12, 0, tzinfo=timezone.utc)  # Wednesday
+        >>> end = datetime(2024, 1, 22, 0, 0, tzinfo=timezone.utc)
+        >>> list(get_timestamp_range(start, end, "week"))
+        [datetime(2024, 1, 8, 0, 0, tzinfo=timezone.utc),   # Monday
+         datetime(2024, 1, 15, 0, 0, tzinfo=timezone.utc)]  # Next Monday
+
+    Note:
+        - If end_time <= start_time (after rounding), returns an empty iterator
+        - Week intervals always start on Monday (weekday 0)
+        - Month intervals handle variable month lengths correctly including leap years
+        - The function works in local timezone for stride calculations but returns UTC
+    """
+    if not is_timezone_aware(start_time) or not is_timezone_aware(end_time):
+        raise ValueError("start_time and end_time must be timezone-aware")
+
+    # Apply UTC offset to work in local timezone
+    offset_delta = timedelta(minutes=utc_offset_minutes)
+    local_start_time = start_time + offset_delta
+
+    # round down start_time to the nearest stride in local timezone
+    if stride == "minute":
+        t = local_start_time.replace(second=0, microsecond=0)
+    elif stride == "hour":
+        t = local_start_time.replace(minute=0, second=0, microsecond=0)
+    elif stride == "day":
+        t = local_start_time.replace(hour=0, minute=0, second=0, microsecond=0)
+    elif stride == "week":
+        # Round down to the beginning of the week (Monday)
+        days_since_monday = local_start_time.weekday()
+        t = local_start_time.replace(hour=0, minute=0, second=0, microsecond=0) - timedelta(
+            days=days_since_monday
+        )
+    elif stride == "month":
+        t = local_start_time.replace(day=1, hour=0, minute=0, second=0, microsecond=0)
+    elif stride == "year":
+        t = local_start_time.replace(month=1, day=1, hour=0, minute=0, second=0, microsecond=0)
+    else:
+        assert_never(stride)
+
+    # Convert back to UTC for comparisons and yielding
+    local_end_time = end_time + offset_delta
+
+    while t < local_end_time:
+        # Yield timestamp converted back to UTC
+        yield t - offset_delta
+        if stride == "minute":
+            t += timedelta(minutes=1)
+        elif stride == "hour":
+            t += timedelta(hours=1)
+        elif stride == "day":
+            t += timedelta(days=1)
+        elif stride == "week":
+            t += timedelta(weeks=1)
+        elif stride == "month":
+            next_month = t.month % 12 + 1
+            next_year = t.year + (t.month // 12)
+            t = t.replace(
+                year=next_year, month=next_month, day=1, hour=0, minute=0, second=0, microsecond=0
+            )
+        elif stride == "year":
+            t = t.replace(
+                year=t.year + 1, month=1, day=1, hour=0, minute=0, second=0, microsecond=0
+            )
+        else:
+            assert_never(stride)