Add Python 3.13, Remove Python 3.8 and 3.9

.github/workflows/hamilton-main.yml

@@ -79,15 +79,6 @@ jobs:
 
 
         - name: Test integrations
-          if: ${{ matrix.python-version == '3.9' }}
-          run: |
-            uv sync --group test --extra pandera
-            uv pip install -r tests/integrations/pandera/requirements.txt
-            uv pip install dask-expr
-            uv run pytest tests/integrations
-
-        - name: Test integrations
-          if: ${{ matrix.python-version != '3.9' }}
           run: |
             uv sync --group test --extra pandera
             uv pip install -r tests/integrations/pandera/requirements.txt
@@ -111,24 +102,19 @@ jobs:
             uv run pytest plugin_tests/h_narwhals
 
         - name: Test dask
-          # Dask supports >= py3.9
-          if: ${{ matrix.python-version != '3.8' }}
           run: |
             uv sync --group test --extra dask
             uv run pytest plugin_tests/h_dask
 
         - name: Test ray
-          # Ray supports >= py3.9
-          if: ${{ matrix.python-version != '3.8' }}
           env:
             RAY_ENABLE_UV_RUN_RUNTIME_ENV: 0 # https://github.com/ray-project/ray/issues/53848
           run: |
             uv sync --group test --extra ray
             uv run pytest plugin_tests/h_ray
 
         - name: Test pyspark
-          # Spark supports >= py3.9
-          if: ${{ matrix.python-version != '3.8' && runner.os == 'Linux' }}
+          if: ${{ runner.os == 'Linux' }}
           env:
             PYSPARK_SUBMIT_ARGS: "--conf spark.sql.ansi.enabled=false pyspark-shell"
           run: |
@@ -139,8 +125,7 @@ jobs:
             uv run pytest plugin_tests/h_spark
 
         - name: Test pyspark
-          # Spark supports >= py3.9
-          if: ${{ matrix.python-version != '3.8' && runner.os != 'Linux' }}
+          if: ${{ runner.os != 'Linux' }}
           env:
             PYSPARK_SUBMIT_ARGS: "--conf spark.sql.ansi.enabled=false pyspark-shell"
           run: |
@@ -151,7 +136,7 @@ jobs:
 
         - name: Test vaex
           # Vaex supports <= py3.10 and numpy<2
-          if: ${{ runner.os == 'Linux' && (matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10') }}
+          if: ${{ runner.os == 'Linux' && matrix.python-version == '3.10' }}
           run: |
             sudo apt-get install --no-install-recommends --yes libpcre3-dev cargo
             uv sync --group test --extra vaex
@@ -160,7 +145,7 @@ jobs:
 
         - name: Test vaex
           # Vaex supports <= py3.10 and numpy<2
-          if: ${{ runner.os != 'Linux' && (matrix.python-version == '3.8' || matrix.python-version == '3.9' || matrix.python-version == '3.10') }}
+          if: ${{ runner.os != 'Linux' && matrix.python-version == '3.10' }}
           run: |
             uv sync --group test --extra vaex
             uv pip install "numpy<2"

.github/workflows/hamilton-ui-backend.yml

@@ -42,7 +42,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
-          python-version: '3.9'
+          python-version: '3.10'
       - name: Install dependencies
         run: |
           cd ui/backend/server

contrib/docs/compile_docs.py

@@ -344,7 +344,7 @@ def _create_commit_file(df_path, single_df):
     os.makedirs(commit_path, exist_ok=True)
     with open(os.path.join(commit_path, "commit.txt"), "w") as f:
         for commit, ts in zip(
-            single_df["__init__.py"]["commit"], single_df["__init__.py"]["timestamp"]
+            single_df["__init__.py"]["commit"], single_df["__init__.py"]["timestamp"], strict=False
         ):
             f.write(f"[commit::{commit}][ts::{ts}]\n")
 

contrib/hamilton/contrib/user/skrawcz/customize_embeddings/__init__.py

@@ -409,7 +409,7 @@ def _accuracy_and_se(
         threshold = threshold_thousandths / 1000
         total = 0
         correct = 0
-        for cs, ls in zip(cosine_similarity, labeled_similarity):
+        for cs, ls in zip(cosine_similarity, labeled_similarity, strict=False):
             total += 1
             if cs > threshold:
                 prediction = 1

contrib/hamilton/contrib/user/skrawcz/fine_tuning/__init__.py

@@ -657,7 +657,7 @@ def hold_out_set_predictions(
             )
             predictions.append(prediction)
             questions.append(tokenizer.decode(sample["input_ids"], skip_special_tokens=True))
-    return list(zip(questions, predictions))
+    return list(zip(questions, predictions, strict=False))
 
 
 if __name__ == "__main__":

contrib/setup.py

@@ -75,15 +75,14 @@ def load_requirements():
         "Natural Language :: English",
         "License :: OSI Approved :: Apache Software License",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.8",
-        "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
         "Programming Language :: Python :: 3.12",
+        "Programming Language :: Python :: 3.13",
     ],
     # Note that this feature requires pep8 >= v9 and a version of setup tools greater than the
     # default version installed with virtualenv. Make sure to update your tools!
-    python_requires=">=3.8, <4",
+    python_requires=">=3.10, <4",
     # adding this to slim the package down, since these dependencies are only used in certain contexts.
     extras_require={
         "visualization": ["sf-hamilton[visualization]"],

dev_tools/language_server/pyproject.toml

@@ -26,7 +26,7 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3 :: Only",
 ]
-requires-python = ">=3.8, <4"
+requires-python = ">=3.10, <4"
 dependencies = [
     "pygls>=1.3.1",
     "sf-hamilton[visualization]>=1.56",

docs/get-started/install.md

@@ -24,7 +24,7 @@ Installing hamilton is easy!
 
 Apache Hamilton is a lightweight framework with a variety of extensions/plugins. To get started, you'll need the following:
 
-- ``python >= 3.8``
+- ``python >= 3.10``
 - ``pip``
 
 For help with python/pip/managing virtual environments see the [python docs](https://docs.python.org/3/tutorial/venv.html/).

examples/LLM_Workflows/modular_llm_stack/lancedb_module.py

@@ -68,7 +68,9 @@ def data_objects(
     assert len(ids) == len(titles) == len(text_contents) == len(embeddings)
     return [
         dict(squad_id=id_, title=title, context=context, vector=embedding, **metadata)
-        for id_, title, context, embedding in zip(ids, titles, text_contents, embeddings)
+        for id_, title, context, embedding in zip(
+            ids, titles, text_contents, embeddings, strict=False
+        )
     ]
 
 

examples/LLM_Workflows/modular_llm_stack/marqo_module.py

@@ -58,7 +58,7 @@ def data_objects(
     assert len(ids) == len(titles) == len(text_contents)
     return [
         dict(_id=id, title=title, Description=text_content)
-        for id, title, text_content in zip(ids, titles, text_contents)
+        for id, title, text_content in zip(ids, titles, text_contents, strict=False)
         if id is not None and title is not None or text_content is not None
     ]
 

examples/LLM_Workflows/modular_llm_stack/pinecone_module.py

@@ -62,7 +62,7 @@ def data_objects(
     assert len(ids) == len(titles) == len(embeddings)
     properties = [dict(title=title, **metadata) for title in titles]
     embeddings = [x.tolist() for x in embeddings]
-    return list(zip(ids, embeddings, properties))
+    return list(zip(ids, embeddings, properties, strict=False))
 
 
 def push_to_vector_db(

examples/LLM_Workflows/modular_llm_stack/qdrant_module.py

@@ -59,7 +59,7 @@ def data_objects(
     ids = list(range(len(ids)))
     payloads = [
         dict(id=_id, text_content=text_content, title=title, **metadata)
-        for _id, title, text_content in zip(ids, titles, text_contents)
+        for _id, title, text_content in zip(ids, titles, text_contents, strict=False)
     ]
     embeddings = [x.tolist() for x in embeddings]
     return dict(ids=ids, vectors=embeddings, payload=payloads)

examples/LLM_Workflows/modular_llm_stack/weaviate_module.py

@@ -86,7 +86,7 @@ def data_objects(
     assert len(ids) == len(titles) == len(text_contents)
     return [
         dict(squad_id=id_, title=title, context=context, **metadata)
-        for id_, title, context in zip(ids, titles, text_contents)
+        for id_, title, context in zip(ids, titles, text_contents, strict=False)
     ]
 
 

examples/LLM_Workflows/retrieval_augmented_generation/backend/ingestion.py

@@ -212,7 +212,9 @@ def store_documents(
                 uuid=document_uuid,
             )
 
-            chunk_iterator = zip(pdf_obj["chunked_text"], pdf_obj["chunked_embeddings"])
+            chunk_iterator = zip(
+                pdf_obj["chunked_text"], pdf_obj["chunked_embeddings"], strict=False
+            )
             for chunk_idx, (chunk_text, chunk_embedding) in enumerate(chunk_iterator):
                 chunk_object = dict(content=chunk_text, chunk_index=chunk_idx)
                 chunk_uuid = generate_uuid5(chunk_object, "Chunk")

examples/due_date_probabilities/probability_estimation.py

@@ -133,7 +133,8 @@ def raw_probabilities(raw_data: str) -> pd.DataFrame:
     days = [int(item.split(", ")[1].split()[0]) for item in raw_data]
     probability = [float(item.split()[5].replace("%", "")) / 100 for item in raw_data]
     probabilities_data = [
-        (week * 7 + day, probability) for week, day, probability in zip(weeks, days, probability)
+        (week * 7 + day, probability)
+        for week, day, probability in zip(weeks, days, probability, strict=False)
     ]
     probabilities_df = pd.DataFrame(probabilities_data)
     probabilities_df.columns = ["days", "probability"]

hamilton/async_driver.py

@@ -38,7 +38,7 @@ async def await_dict_of_tasks(task_dict: Dict[str, typing.Awaitable]) -> Dict[st
     keys = sorted(task_dict.keys())
     coroutines = [task_dict[key] for key in keys]
     coroutines_gathered = await asyncio.gather(*coroutines)
-    return dict(zip(keys, coroutines_gathered))
+    return dict(zip(keys, coroutines_gathered, strict=False))
 
 
 async def process_value(val: Any) -> Any:

hamilton/experimental/decorators/parameterize_frame.py

@@ -67,11 +67,11 @@ def _convert_params_from_df(parameterization: pd.DataFrame) -> List[Parameterize
     for _, column_set in parameterization.iterrows():
         parameterization = {
             arg: dep_type(col_value)
-            for arg, col_value, dep_type in zip(args, column_set, dep_types_converted)
+            for arg, col_value, dep_type in zip(args, column_set, dep_types_converted, strict=False)
             if dep_type is not None
         }
         extracted_columns = [
-            col for col, dep_type in zip(column_set, dep_types) if dep_type == "out"
+            col for col, dep_type in zip(column_set, dep_types, strict=False) if dep_type == "out"
         ]
         out.append(ParameterizedExtract(tuple(extracted_columns), parameterization))
     return out

hamilton/function_modifiers/expanders.py

@@ -1065,7 +1065,9 @@ def tuple_generator(*args, **kwargs):
 
         output_nodes = [node_.copy_with(callabl=tuple_generator)]
 
-        for idx, (field_name, field_type) in enumerate(zip(self.fields, self.field_types)):
+        for idx, (field_name, field_type) in enumerate(
+            zip(self.fields, self.field_types, strict=False)
+        ):
 
             def extractor(field_index: int = idx, **kwargs) -> field_type:  # type: ignore
                 # This extractor is constructed to avoid closure issues.

hamilton/plugins/h_kedro.py

@@ -35,7 +35,7 @@ def expand_k_node(base_node: HNode, outputs: List[str]) -> List[HNode]:
     """
 
     def _convert_output_from_tuple_to_dict(node_result: Any, node_kwargs: Dict[str, Any]):
-        return {out: v for out, v in zip(outputs, node_result)}
+        return {out: v for out, v in zip(outputs, node_result, strict=False)}
 
     # NOTE isinstance(Any, type) is False for Python < 3.11
     extractor = extract_fields(fields={out: Any for out in outputs})
@@ -91,7 +91,7 @@ def k_node_to_h_nodes(node: KNode) -> List[HNode]:
 
     # remap the function parameters to the node `inputs` and clean Kedro `parameters` name
     new_params = {}
-    for param, k_input in zip(params, node.inputs):
+    for param, k_input in zip(params, node.inputs, strict=False):
         if k_input.startswith("params:"):
             k_input = k_input.partition("params:")[-1]
 

hamilton/plugins/h_openlineage.py

@@ -61,6 +61,7 @@ def extract_schema_facet(metadata):
             for k, v in zip(
                 metadata["dataframe_metadata"]["column_names"],
                 metadata["dataframe_metadata"]["datatypes"],
+                strict=False,
             )
         ]
         schema_facet = facet_v2.schema_dataset.SchemaDatasetFacet(

pyproject.toml

@@ -93,10 +93,9 @@ dev = [
   "ruff==0.5.7", # this should match `.pre-commit-config.yaml`
 ]
 test = [
-  "connectorx<=0.3.2; python_version=='3.8'",
-  "connectorx; python_version!='3.8'",
+  "connectorx",
   "dask[complete]",
-  "dask-expr>=1.1.14; python_version >= '3.10'", # Bugfix only available after py3.10 https://github.com/dask/dask-expr/pull/1150
+  "dask-expr>=1.1.14",
   "datasets>=2.18.0", # huggingface datasets -- https://github.com/huggingface/datasets/issues/6737#issuecomment-2107336816
   "diskcache",
   "dlt",
@@ -117,15 +116,13 @@ test = [
   "polars",
   "pyarrow",
   "pydantic >=2.0",
-  "pyreadstat<1.2.8; python_version <= '3.9'", # for SPSS data loader
-  "pyreadstat; python_version > '3.9'", # for SPSS data loader
+  "pyreadstat", # for SPSS data loader
   "pytest",
   "pytest-asyncio",
   "pytest-cov",
   "PyYAML",
   "scikit-learn",
-  "sqlalchemy==1.4.49; python_version == '3.7.*'",
-  "sqlalchemy; python_version >= '3.8'",
+  "sqlalchemy",
   "typer",
   "xgboost",
   "xlsx2csv", # for excel data loader
@@ -135,7 +132,7 @@ docs = [
   {include-group = "dev"},
   "alabaster>=0.7,<0.8,!=0.7.5", # read the docs pins
   "commonmark==0.9.1", # read the docs pins
-  "dask-expr>=1.1.14; python_version >= '3.10'", # Bugfix only available after py3.10 https://github.com/dask/dask-expr/pull/1150
+  "dask-expr>=1.1.14",
   "dask[distributed]",
   "ddtrace<3.0",
   "diskcache",
@@ -200,7 +197,7 @@ slack = "https://join.slack.com/t/hamilton-opensource/shared_invite/zt-2niepkra8
 
 [tool.ruff]
 line-length = 100
-target-version = "py38"  # Must include only the earliest supported version
+target-version = "py310"  # Must include only the earliest supported version
 
 [tool.ruff.format]
 docstring-code-format = false

scripts/build_conda.sh

@@ -22,7 +22,7 @@ CONDA_HOME=$HOME/anaconda3
 # conda activate && anaconda login
 pkg='sf-hamilton'
 # adjust the Python versions you would like to build
-array=(3.7 3.8 3.9 3.10 3.11 )
+array=(3.10 3.11 3.12 3.13)
 echo "Building conda package ..."
 cd ~
 # this will create a ~/sf-hamilton directory with metadata to build the package.

scripts/update_blogs_in_learning_resources.py

@@ -105,7 +105,7 @@ def fetch_articles(url, cutoff_date):
             None,
         ),
     ]
-    for _i, (anchor, time_el) in enumerate(zip(anchors, timeEls)):
+    for _i, (anchor, time_el) in enumerate(zip(anchors, timeEls, strict=False)):
         link = anchor["href"]
         text = anchor.get_text()
 

tests/resources/test_driver_serde_mapper.py

@@ -27,7 +27,7 @@ def mapper(
 ) -> Parallelizable[dict]:
     if final_vars is None:
         final_vars = []
-    for dr, input_ in zip(drivers, inputs):
+    for dr, input_ in zip(drivers, inputs, strict=False):
         yield {
             "dr": dr,
             "final_vars": final_vars or dr.list_available_variables(),

ui/backend/server/trackingserver_run_tracking/api.py

@@ -426,7 +426,7 @@ async def get_latest_template_runs(
     return CatalogZoomResponse(
         node_runs=[
             NodeRunOutWithExtraData.from_orm(node_run, dag_template_id=dag_template_id)
-            for (node_run, dag_template_id) in zip(node_runs, all_dag_versions)
+            for (node_run, dag_template_id) in zip(node_runs, all_dag_versions, strict=False)
         ],
         node_templates=[
             NodeTemplateOut.from_orm(node_template) for node_template in node_templates

ui/backend/setup.py

@@ -54,15 +54,14 @@ def load_requirements():
         "Natural Language :: English",
         "License :: OSI Approved :: Apache Software License",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.8",
-        "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
         "Programming Language :: Python :: 3.12",
+        "Programming Language :: Python :: 3.13",
     ],
     # Note that this feature requires pep8 >= v9 and a version of setup tools greater than the
     # default version installed with virtualenv. Make sure to update your tools!
-    python_requires=">=3.6, <4",
+    python_requires=">=3.10, <4",
     # adding this to slim the package down, since these dependencies are only used in certain contexts.
     # Relevant project URLs
     project_urls={  # Optional

ui/sdk/pyproject.toml

@@ -17,12 +17,12 @@ classifiers = [
     "Natural Language :: English",
     "License :: OSI Approved :: Apache Software License",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.8",
-    "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
-    "Programming Language :: Python :: 3.11"
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13"
 ]
-requires-python = ">=3.7, <4"
+requires-python = ">=3.10, <4"
 dynamic = ["dependencies", "optional-dependencies", "version"]
 
 [project.urls]

writeups/developer_setup.md

@@ -121,7 +121,7 @@ The following values for `TASK` are recognized:
 Choose a Python version and task.
 
 ```shell
-PYTHON_VERSION='3.8'
+PYTHON_VERSION='3.10'
 TASK=tests
 ```