arize-phoenix 5.1.0__tar.gz → 5.1.2__tar.gz

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: arize-phoenix
-Version: 5.1.0
+Version: 5.1.2
 Summary: AI Observability and Evaluation
 Project-URL: Documentation, https://docs.arize.com/phoenix/
 Project-URL: Issues, https://github.com/Arize-ai/phoenix/issues
@@ -27,10 +27,11 @@ Requires-Dist: fastapi
 Requires-Dist: fastapi-mail
 Requires-Dist: grpc-interceptor
 Requires-Dist: grpcio
-Requires-Dist: hdbscan>=0.8.33
+Requires-Dist: hdbscan>=0.8.38
 Requires-Dist: httpx
 Requires-Dist: jinja2
-Requires-Dist: numpy<2
+Requires-Dist: numba>=0.60.0
+Requires-Dist: numpy!=2.0.0
 Requires-Dist: openinference-instrumentation>=0.1.12
 Requires-Dist: openinference-semantic-conventions>=0.1.9
 Requires-Dist: opentelemetry-exporter-otlp

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/pyproject.toml

@@ -22,11 +22,12 @@ classifiers = [
 ]
 dependencies = [
   "scikit-learn",
-  "numpy<2",  # https://github.com/scikit-learn-contrib/hdbscan/issues/642
+  "numpy!=2.0.0",  # https://github.com/lmcinnes/umap/issues/1138
   "pandas>=1.0",
   "jinja2",
   "umap-learn",
-  "hdbscan>=0.8.33",
+  "hdbscan>=0.8.38",
+  "numba>=0.60.0",  # https://github.com/astral-sh/uv/issues/6281
   "starlette",
   "uvicorn",
   "psutil",
@@ -145,15 +146,32 @@ build-backend = "hatchling.build"
 
 [tool.hatch.build.targets.wheel]
 packages = ["src/phoenix"]
-exclude = ["src/phoenix/evals/"]
-artifacts = ["src/phoenix/server/static", "src/phoenix/db/migrations"]
+exclude = [
+  "src/phoenix/evals/",
+  "src/phoenix/otel/",
+]
+artifacts = [
+  "src/phoenix/server/static",
+  "src/phoenix/db/migrations",
+]
 
 [tool.hatch.build]
 only-packages = true
 
 [tool.hatch.build.targets.sdist]
-exclude = ["src/phoenix/evals/", "packages/"]
-artifacts = ["src/phoenix/server/static", "src/phoenix/db/migrations"]
+exclude = [
+  "packages/",
+  "src/phoenix/evals/",
+  "src/phoenix/otel/",
+  "tests/",
+  "integration_tests/",
+  "scripts/",
+  "examples/",
+]
+artifacts = [
+  "src/phoenix/server/static",
+  "src/phoenix/db/migrations",
+]
 
 [tool.hatch.envs.default]
 dependencies = [

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/src/phoenix/pointcloud/clustering.py

@@ -3,7 +3,6 @@ from typing import List, Set
 
 import numpy as np
 import numpy.typing as npt
-from hdbscan import HDBSCAN
 from typing_extensions import TypeAlias
 
 RowIndex: TypeAlias = int
@@ -18,6 +17,8 @@ class Hdbscan:
     cluster_selection_epsilon: float = 0.0
 
     def find_clusters(self, mat: Matrix) -> List[RawCluster]:
+        from hdbscan import HDBSCAN
+
         cluster_ids: npt.NDArray[np.int_] = HDBSCAN(**asdict(self)).fit_predict(mat)
         ans: List[RawCluster] = [set() for _ in range(np.max(cluster_ids) + 1)]
         for row_idx, cluster_id in enumerate(cluster_ids):

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/src/phoenix/pointcloud/projectors.py

@@ -6,12 +6,6 @@ import numpy as np
 import numpy.typing as npt
 from typing_extensions import TypeAlias
 
-with warnings.catch_warnings():
-    from numba.core.errors import NumbaWarning
-
-    warnings.simplefilter("ignore", category=NumbaWarning)
-    from umap import UMAP
-
 Matrix: TypeAlias = npt.NDArray[np.float64]
 
 
@@ -25,6 +19,11 @@ class Umap:
     min_dist: float = 0.1
 
     def project(self, mat: Matrix, n_components: int) -> Matrix:
+        with warnings.catch_warnings():
+            from numba.core.errors import NumbaWarning
+
+            warnings.simplefilter("ignore", category=NumbaWarning)
+            from umap import UMAP
         config = asdict(self)
         config["n_components"] = n_components
         if len(mat) <= n_components:

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/src/phoenix/server/api/types/EmbeddingDimension.py

@@ -422,17 +422,17 @@ class EmbeddingDimension(Node):
             if isinstance(inferences_role, InferencesRole):
                 dataset = model[inferences_role.value]
                 embedding_metadata = EmbeddingMetadata(
-                    prediction_id=dataset[PREDICTION_ID][row_id],
-                    link_to_data=dataset[self.dimension.link_to_data][row_id],
-                    raw_data=dataset[self.dimension.raw_data][row_id],
+                    prediction_id=dataset[PREDICTION_ID].iloc[row_id],
+                    link_to_data=dataset[self.dimension.link_to_data].iloc[row_id],
+                    raw_data=dataset[self.dimension.raw_data].iloc[row_id],
                 )
             elif (corpus := info.context.corpus) is not None:
                 dataset = corpus[PRIMARY]
                 dimension = cast(ms.EmbeddingDimension, corpus[PROMPT])
                 embedding_metadata = EmbeddingMetadata(
-                    prediction_id=dataset[PREDICTION_ID][row_id],
-                    link_to_data=dataset[dimension.link_to_data][row_id],
-                    raw_data=dataset[dimension.raw_data][row_id],
+                    prediction_id=dataset[PREDICTION_ID].iloc[row_id],
+                    link_to_data=dataset[dimension.link_to_data].iloc[row_id],
+                    raw_data=dataset[dimension.raw_data].iloc[row_id],
                 )
             else:
                 continue
@@ -445,10 +445,10 @@ class EmbeddingDimension(Node):
                     event_id=event_id,
                     coordinates=to_gql_coordinates(vector),
                     event_metadata=EventMetadata(
-                        prediction_label=dataset[PREDICTION_LABEL][row_id],
-                        prediction_score=dataset[PREDICTION_SCORE][row_id],
-                        actual_label=dataset[ACTUAL_LABEL][row_id],
-                        actual_score=dataset[ACTUAL_SCORE][row_id],
+                        prediction_label=dataset[PREDICTION_LABEL].iloc[row_id],
+                        prediction_score=dataset[PREDICTION_SCORE].iloc[row_id],
+                        actual_label=dataset[ACTUAL_LABEL].iloc[row_id],
+                        actual_score=dataset[ACTUAL_SCORE].iloc[row_id],
                     ),
                     embedding_metadata=embedding_metadata,
                 )

{arize_phoenix-5.1.0 → arize_phoenix-5.1.2}/src/phoenix/trace/fixtures.py

@@ -147,21 +147,21 @@ demo_llama_index_rag_fixture = TracesFixture(
 
 demo_code_based_agent_fixture = TracesFixture(
     name="demo_code_based_agent",
-    project_name="code_based_agent_demo",
+    project_name="demo_code_based_agent",
     description="Shows traces captured from a code-based agent.",
     file_name="agent-demo-traces.parquet",
 )
 
 demo_langgraph_agent_fixture = TracesFixture(
     name="demo_langgraph_agent",
-    project_name="langgraph_agent_demo",
+    project_name="demo_langgraph_agent",
     description="Shows traces captured from a Langgraph agent.",
-    file_name="langgraph-demo-traces.parquet",
+    file_name="langgraph-demo-traces-format-updated.parquet",
 )
 
 demo_llamaindex_workflows_agent_fixture = TracesFixture(
     name="demo_llamaindex_workflows_agent",
-    project_name="llamaindex_workflows_agent_demo",
+    project_name="demo_llamaindex_workflows_agent",
     description="Shows traces captured from a LlamaIndex Workflows agent.",
     file_name="llamaindex-workflow-demo-traces.parquet",
 )

arize_phoenix-5.1.2/src/phoenix/version.py

@@ -0,0 +1 @@
+__version__ = "5.1.2"

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/chat-service/chat/app.py

@@ -1,227 +0,0 @@
-import json
-import os
-import warnings
-from typing import Any, Dict, Iterator, List, Tuple
-from uuid import uuid4
-
-from fastapi import FastAPI, HTTPException
-from httpx import AsyncClient, ConnectError
-from openinference.semconv.trace import (
-    MessageAttributes,
-    OpenInferenceMimeTypeValues,
-    OpenInferenceSpanKindValues,
-    SpanAttributes,
-)
-from opentelemetry import trace as trace_api
-from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
-from opentelemetry.sdk import trace as trace_sdk
-from opentelemetry.sdk.trace.export import (
-    SimpleSpanProcessor,
-)
-
-from chat.types import FeedbackRequest, Message, MessagesPayload, MessagesResponse
-
-COLLECTOR_HOST = os.getenv("COLLECTOR_HOST", "localhost")
-endpoint = f"http://{COLLECTOR_HOST}:6006/v1"
-tracer_provider = trace_sdk.TracerProvider()
-tracer_provider.add_span_processor(
-    SimpleSpanProcessor(OTLPSpanExporter(f"{endpoint}/traces"))
-)
-trace_api.set_tracer_provider(tracer_provider)
-tracer = trace_api.get_tracer(__name__)
-
-
-def getenv_or_raise(key: str) -> str:
-    if not (value := os.getenv(key)):
-        raise ValueError(f"Please set the {key} environment variable.")
-    return value
-
-
-OPENAI_API_KEY = getenv_or_raise("OPENAI_API_KEY")
-OPENAI_API_URL = "https://api.openai.com/v1/chat/completions"
-OPENAI_MODEL = "gpt-4"
-
-http_client = AsyncClient()
-app = FastAPI()
-
-
-class OpenAIException(HTTPException):
-    pass
-
-
-@app.post("/messages/")
-async def messages(messages_payload: MessagesPayload) -> MessagesResponse:
-    messages = messages_payload.messages
-    invocation_parameters = {"temperature": 0.1}
-    openai_payload = {
-        "model": OPENAI_MODEL,
-        **invocation_parameters,
-        "messages": [message.model_dump() for message in messages],
-    }
-    with tracer.start_as_current_span("OpenAI Async Chat Completion") as span:
-        for attribute_key, attribute_value in (
-            *_llm_span_kind_attributes(),
-            *_llm_model_name_attributes(OPENAI_MODEL),
-            *_llm_invocation_parameters_attributes(invocation_parameters),
-            *_input_attributes(openai_payload),
-            *_llm_input_messages_attributes(messages),
-        ):
-            span.set_attribute(attribute_key, attribute_value)
-        response = await http_client.post(
-            OPENAI_API_URL,
-            headers={
-                "Content-Type": "application/json",
-                "Authorization": f"Bearer {OPENAI_API_KEY}",
-            },
-            json=openai_payload,
-        )
-        if not (200 <= response.status_code < 300):
-            raise OpenAIException(
-                status_code=500, detail=response.content.decode("utf-8")
-            )
-        span.set_status(trace_api.StatusCode.OK)
-        response_data = response.json()
-        assistant_message_content = response_data["choices"][0]["message"]["content"]
-        message_uuid = str(uuid4())
-        assistant_message = Message(
-            role="assistant",
-            content=assistant_message_content,
-            uuid=message_uuid,
-        )
-        for (
-            attribute_key,
-            attribute_value,
-        ) in (
-            *_output_attributes(response_data),
-            *_llm_output_message_attributes(assistant_message),
-            *_llm_token_usage_attributes(response_data),
-        ):
-            span.set_attribute(attribute_key, attribute_value)
-        span_id = span.get_span_context().span_id.to_bytes(8, "big").hex()
-        assistant_message.span_id = span_id
-
-    return MessagesResponse(message=assistant_message)
-
-
-@app.post("/feedback/")
-async def post_feedback(feedback_request: FeedbackRequest) -> None:
-    if feedback_request.feedback == 1:
-        label = "👍"
-    elif feedback_request.feedback == 0:
-        label = "👎"
-
-    request_body = {
-        "data": [
-            {
-                "span_id": feedback_request.span_id,
-                "name": "user_feedback",
-                "annotator_kind": "HUMAN",
-                "result": {"label": label, "score": feedback_request.feedback},
-                "metadata": {},
-            }
-        ]
-    }
-
-    try:
-        await http_client.post(
-            f"{endpoint}/span_annotations",
-            json=request_body,
-        )
-    except ConnectError:
-        warnings.warn("Could not connect to Phoenix server.")
-
-
-def _llm_span_kind_attributes() -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference span kind attribute for LLMs.
-    """
-    yield SpanAttributes.OPENINFERENCE_SPAN_KIND, OpenInferenceSpanKindValues.LLM.value
-
-
-def _llm_model_name_attributes(model_name: str) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference model name attribute.
-    """
-    yield SpanAttributes.LLM_MODEL_NAME, model_name
-
-
-def _llm_invocation_parameters_attributes(
-    invocation_parameters: Dict[str, Any],
-) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference invocation parameters attribute as a JSON string.
-    """
-    yield SpanAttributes.LLM_INVOCATION_PARAMETERS, json.dumps(invocation_parameters)
-
-
-def _input_attributes(payload: Any) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference input value attribute as a JSON string if the
-    payload can be serialized as JSON, otherwise as a string.
-    """
-    try:
-        yield SpanAttributes.INPUT_VALUE, json.dumps(payload)
-        yield SpanAttributes.INPUT_MIME_TYPE, OpenInferenceMimeTypeValues.JSON.value
-    except json.JSONDecodeError:
-        yield SpanAttributes.INPUT_VALUE, str(payload)
-        yield SpanAttributes.INPUT_MIME_TYPE, OpenInferenceMimeTypeValues.TEXT.value
-
-
-def _llm_input_messages_attributes(
-    messages: List[Message],
-) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference input messages attributes for each message in the list.
-    """
-    for messages_index, message in enumerate(messages):
-        yield (
-            f"{SpanAttributes.LLM_INPUT_MESSAGES}.{messages_index}.{MessageAttributes.MESSAGE_ROLE}",
-            message.role,
-        )
-        yield (
-            f"{SpanAttributes.LLM_INPUT_MESSAGES}.{messages_index}.{MessageAttributes.MESSAGE_CONTENT}",
-            message.content,
-        )
-
-
-def _output_attributes(payload: Any) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference output value attribute as a JSON string if the
-    payload can be serialized as JSON, otherwise as a string.
-    """
-    try:
-        yield SpanAttributes.OUTPUT_VALUE, json.dumps(payload)
-        yield SpanAttributes.OUTPUT_MIME_TYPE, OpenInferenceMimeTypeValues.JSON.value
-    except json.JSONDecodeError:
-        yield SpanAttributes.OUTPUT_VALUE, str(payload)
-        yield SpanAttributes.OUTPUT_MIME_TYPE, OpenInferenceMimeTypeValues.TEXT.value
-
-
-def _llm_output_message_attributes(message: Message) -> Iterator[Tuple[str, str]]:
-    """
-    Yields the OpenInference output message attributes.
-    """
-    yield (
-        f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_ROLE}",
-        message.role,
-    )
-    yield (
-        f"{SpanAttributes.LLM_OUTPUT_MESSAGES}.0.{MessageAttributes.MESSAGE_CONTENT}",
-        message.content,
-    )
-
-
-def _llm_token_usage_attributes(
-    response_data: Dict[str, Any],
-) -> Iterator[Tuple[str, int]]:
-    """
-    Parses and yields token usage attributes from the response data.
-    """
-    if not isinstance((usage := response_data.get("usage")), dict):
-        return
-    if prompt_tokens := usage.get("prompt_tokens"):
-        yield SpanAttributes.LLM_TOKEN_COUNT_PROMPT, prompt_tokens
-    if completion_tokens := usage.get("completion_tokens"):
-        yield SpanAttributes.LLM_TOKEN_COUNT_COMPLETION, completion_tokens
-    if total_tokens := usage.get("total_tokens"):
-        yield SpanAttributes.LLM_TOKEN_COUNT_TOTAL, total_tokens

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/chat-service/chat/types.py

@@ -1,25 +0,0 @@
-from typing import List, Literal, Optional
-
-from pydantic import BaseModel
-
-Role = Literal["system", "assistant", "user"]
-
-
-class Message(BaseModel):
-    role: Role
-    content: str
-    uuid: str
-    span_id: Optional[str] = None
-
-
-class MessagesPayload(BaseModel):
-    messages: List[Message]
-
-
-class MessagesResponse(BaseModel):
-    message: Message
-
-
-class FeedbackRequest(BaseModel):
-    feedback: int
-    span_id: str

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/frontend/Dockerfile

@@ -1,15 +0,0 @@
-FROM python:3.11-slim as builder
-WORKDIR /app
-COPY requirements.txt /app/
-RUN pip install uv
-RUN uv venv
-ENV PATH="/app/.venv/bin:$PATH"
-RUN uv pip sync requirements.txt
-
-FROM gcr.io/distroless/python3-debian12
-WORKDIR /app
-COPY --from=builder /app/.venv/lib/python3.11/site-packages/ /app/site-packages/
-ENV PYTHONPATH="/app/site-packages:$PYTHONPATH"
-COPY src/ /app/src/
-EXPOSE 8501
-CMD ["-m", "streamlit", "run", "src/app.py", "--server.port=8501", "--server.address=0.0.0.0"]

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/frontend/Makefile

@@ -1,20 +0,0 @@
-frontend:
-	streamlit run ./src/app.py
-
-format:
-	ruff check --fix . && ruff format .
-
-typecheck:
-	pyright .
-
-types:
-	datamodel-codegen  --input schema.json --input-file-type openapi --output frontend/request_types.py --enum-field-as-literal all
-
-requirements:
-	uv pip compile pyproject.toml > requirements.txt
-
-docker-build:
-	docker build -t frontend .
-
-docker-run:
-	docker run -p 8501:8501 -it frontend

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/frontend/pyproject.toml

@@ -1,24 +0,0 @@
-[project]
-name = "frontend"
-version = "0.1.0"
-dependencies = [
-  "streamlit",
-  "httpx",
-  "pydantic",
-]
-requires-python = ">=3.8"
-
-[project.optional-dependencies]
-dev = [
-  "pyright",
-  "ruff",
-  "datamodel-code-generator",
-]
-
-[tool.ruff]
-exclude = [
-  "frontend/request_types.py",
-]
-
-[tool.ruff.lint]
-extend-select = ["I"]

arize_phoenix-5.1.0/examples/manually-instrumented-chatbot/frontend/requirements.txt

@@ -1,123 +0,0 @@
-# This file was autogenerated by uv via the following command:
-#    uv pip compile pyproject.toml
-altair==5.3.0
-    # via streamlit
-annotated-types==0.6.0
-    # via pydantic
-anyio==4.3.0
-    # via httpx
-attrs==23.2.0
-    # via
-    #   jsonschema
-    #   referencing
-blinker==1.7.0
-    # via streamlit
-cachetools==5.3.3
-    # via streamlit
-certifi==2024.7.4
-    # via
-    #   httpcore
-    #   httpx
-    #   requests
-charset-normalizer==3.3.2
-    # via requests
-click==8.1.7
-    # via streamlit
-gitdb==4.0.11
-    # via gitpython
-gitpython==3.1.43
-    # via streamlit
-h11==0.14.0
-    # via httpcore
-httpcore==1.0.5
-    # via httpx
-httpx==0.27.0
-idna==3.6
-    # via
-    #   anyio
-    #   httpx
-    #   requests
-jinja2==3.1.3
-    # via
-    #   altair
-    #   pydeck
-jsonschema==4.21.1
-    # via altair
-jsonschema-specifications==2023.12.1
-    # via jsonschema
-markdown-it-py==3.0.0
-    # via rich
-markupsafe==2.1.5
-    # via jinja2
-mdurl==0.1.2
-    # via markdown-it-py
-numpy==1.26.4
-    # via
-    #   altair
-    #   pandas
-    #   pyarrow
-    #   pydeck
-    #   streamlit
-packaging==24.0
-    # via
-    #   altair
-    #   streamlit
-pandas==2.2.1
-    # via
-    #   altair
-    #   streamlit
-pillow==10.3.0
-    # via streamlit
-protobuf==4.25.3
-    # via streamlit
-pyarrow==15.0.2
-    # via streamlit
-pydantic==2.6.4
-pydantic-core==2.16.3
-    # via pydantic
-pydeck==0.8.0
-    # via streamlit
-pygments==2.17.2
-    # via rich
-python-dateutil==2.9.0.post0
-    # via pandas
-pytz==2024.1
-    # via pandas
-referencing==0.34.0
-    # via
-    #   jsonschema
-    #   jsonschema-specifications
-requests==2.32.2
-    # via streamlit
-rich==13.7.1
-    # via streamlit
-rpds-py==0.18.0
-    # via
-    #   jsonschema
-    #   referencing
-six==1.16.0
-    # via python-dateutil
-smmap==5.0.1
-    # via gitdb
-sniffio==1.3.1
-    # via
-    #   anyio
-    #   httpx
-streamlit==1.37.0
-tenacity==8.2.3
-    # via streamlit
-toml==0.10.2
-    # via streamlit
-toolz==0.12.1
-    # via altair
-tornado==6.4.1
-    # via streamlit
-typing-extensions==4.11.0
-    # via
-    #   pydantic
-    #   pydantic-core
-    #   streamlit
-tzdata==2024.1
-    # via pandas
-urllib3==2.2.1
-    # via requests