Add ruff rules for naming (#545)

Author: cbornet

docs/modules/examples/pages/qa-with-cassio.adoc

@@ -113,7 +113,7 @@ for filename in SAMPLEDATA:
 +
 [source,python]
 ----
-cassVStore = Cassandra.from_documents(
+cass_vstore = Cassandra.from_documents(
     documents=documents,
     embedding=OpenAIEmbeddings(),
     session=session,
@@ -137,7 +137,7 @@ print(f"\nProcessing done.")
 ----
 prompt = "Who is Luchesi?"
 
-matched_docs = cassVStore.similarity_search(query=prompt, k=1)
+matched_docs = cass_vstore.similarity_search(query=prompt, k=1)
 
 for i, d in enumerate(matched_docs):
     print(f"\n## Document {i}\n")
@@ -149,7 +149,7 @@ Create an Index on top of the vector store.
 +
 [source,python]
 ----
-index = VectorStoreIndexWrapper(vectorstore=cassVStore)
+index = VectorStoreIndexWrapper(vectorstore=cass_vstore)
 ----
 +
 . Create a retriever from the Index.
@@ -182,7 +182,7 @@ Your answer:
 """
 prompt = ChatPromptTemplate.from_template(prompt)
 
-qa = RetrievalQA.from_chain_type(llm=OpenAI(), retriever=cassVStore.as_retriever(), chain_type_kwargs={"prompt": prompt})
+qa = RetrievalQA.from_chain_type(llm=OpenAI(), retriever=cass_vstore.as_retriever(), chain_type_kwargs={"prompt": prompt})
 
 result = qa.run("{question: Who is Luchesi?")
 result

docs/modules/examples/pages/rag-with-cassio.adoc

@@ -142,7 +142,7 @@ for filename in SAMPLEDATA:
   if len(new_docs) > 0:
     documents.extend(new_docs)
 
-cassVStore = Cassandra.from_documents(
+cass_vstore = Cassandra.from_documents(
   documents=documents,
   embedding=OpenAIEmbeddings(),
   session=session,
@@ -158,7 +158,7 @@ print(f"\nProcessing done.")
 Use `VectorStoreIndexWrapper` from `langchain.indexes.vectorstore` for querying.
 [source,python]
 ----
-index = VectorStoreIndexWrapper(vectorstore=cassVStore)
+index = VectorStoreIndexWrapper(vectorstore=cass_vstore)
 query = "Who is Luchesi?"
 index.query(query,llm=llm)
 query = "What motivates Montresor to seek revenge against Fortunato?"

examples/evaluation/langchain_trulens_full.ipynb

@@ -336,12 +336,12 @@
     "from trulens_eval.feedback.provider import AzureOpenAI\n",
     "\n",
     "# Initialize provider class\n",
-    "azureOpenAI = AzureOpenAI(deployment_name=\"gpt-35-turbo\")\n",
+    "azure_open_ai = AzureOpenAI(deployment_name=\"gpt-35-turbo\")\n",
     "\n",
     "context = App.select_context(rag_chain)\n",
     "\n",
     "# Define a groundedness feedback function\n",
-    "grounded = Groundedness(groundedness_provider=azureOpenAI)\n",
+    "grounded = Groundedness(groundedness_provider=azure_open_ai)\n",
     "f_groundedness = (\n",
     "    Feedback(grounded.groundedness_measure_with_cot_reasons, name=\"groundedness\")\n",
     "    .on(context.collect())\n",
@@ -351,19 +351,19 @@
     "\n",
     "# Question/answer relevance between overall question and answer.\n",
     "f_answer_relevance = Feedback(\n",
-    "    azureOpenAI.relevance_with_cot_reasons, name=\"answer_relevance\"\n",
+    "    azure_open_ai.relevance_with_cot_reasons, name=\"answer_relevance\"\n",
     ").on_input_output()\n",
     "\n",
     "# Question/statement relevance between question and each context chunk.\n",
     "f_context_relevance = (\n",
-    "    Feedback(azureOpenAI.qs_relevance_with_cot_reasons, name=\"context_relevance\")\n",
+    "    Feedback(azure_open_ai.qs_relevance_with_cot_reasons, name=\"context_relevance\")\n",
     "    .on_input()\n",
     "    .on(context)\n",
     "    .aggregate(np.mean)\n",
     ")\n",
     "\n",
     "# GroundTruth for comparing the Answer to the Ground-Truth Answer\n",
-    "ground_truth_collection = GroundTruthAgreement(golden_set, provider=azureOpenAI)\n",
+    "ground_truth_collection = GroundTruthAgreement(golden_set, provider=azure_open_ai)\n",
     "f_answer_correctness = Feedback(\n",
     "    ground_truth_collection.agreement_measure, name=\"answer_correctness\"\n",
     ").on_input_output()"
@@ -441,7 +441,7 @@
    "outputs": [],
    "source": [
     "# this downloads the full set of records from the database for an app(s)\n",
-    "dfRecords, feedbackColumns = tru.get_records_and_feedback([app_id])"
+    "df_records, feedback_columns = tru.get_records_and_feedback([app_id])"
    ]
   },
   {
@@ -455,7 +455,7 @@
     "\n",
     "# note that token & cost data collection is currently broken with AzureOpenAI\n",
     "\n",
-    "columns_to_keep = feedbackColumns + [\n",
+    "columns_to_keep = feedback_columns + [\n",
     "    \"record_id\",\n",
     "    \"input\",\n",
     "    \"output\",\n",
@@ -465,11 +465,11 @@
     "    \"total_cost\",\n",
     "]\n",
     "\n",
-    "columns_to_drop = [col for col in dfRecords.columns if col not in columns_to_keep]\n",
+    "columns_to_drop = [col for col in df_records.columns if col not in columns_to_keep]\n",
     "\n",
-    "dfRecords.drop(columns=columns_to_drop, inplace=True)\n",
+    "df_records.drop(columns=columns_to_drop, inplace=True)\n",
     "\n",
-    "dfRecords"
+    "df_records"
    ]
   },
   {
@@ -490,7 +490,7 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "\n",
-    "tests = feedbackColumns + [\"latency\", \"total_tokens\", \"total_cost\"]\n",
+    "tests = feedback_columns + [\"latency\", \"total_tokens\", \"total_cost\"]\n",
     "\n",
     "results = pd.DataFrame(\n",
     "    columns=[\"records\", \"mean\", \"median\", \"95th_percentile\", \"99th_percentile\"]\n",
@@ -506,7 +506,7 @@
     ")\n",
     "\n",
     "for test in tests:\n",
-    "    data = dfRecords[test].dropna().to_list()\n",
+    "    data = df_records[test].dropna().to_list()\n",
     "\n",
     "    records = len(data)\n",
     "    mean = np.mean(data)\n",

examples/evaluation/tru_download.py

@@ -3,14 +3,14 @@
 
 tru = tru_shared.init_tru()
 
-dfAll = pd.DataFrame()
+df_all = pd.DataFrame()
 
 for app in tru.get_apps():
     app_id = app["app_id"]
     print(f"Downloading data for {app_id}...")
-    dfRecords, feedbackColumns = tru.get_records_and_feedback([app_id])
+    df_records, feedback_columns = tru.get_records_and_feedback([app_id])
 
-    columns_to_keep = feedbackColumns + [
+    columns_to_keep = feedback_columns + [
         "record_id",
         "input",
         "output",
@@ -19,16 +19,16 @@
         "total_tokens",
         "total_cost",
     ]
-    columns_to_drop = [col for col in dfRecords.columns if col not in columns_to_keep]
+    columns_to_drop = [col for col in df_records.columns if col not in columns_to_keep]
 
-    dfRecords.drop(columns=columns_to_drop, inplace=True)
+    df_records.drop(columns=columns_to_drop, inplace=True)
 
-    dfRecords["test"] = app_id.split("#")[0]
-    dfRecords["test_uuid"] = app_id.split("#")[1]
-    dfRecords["dataset"] = app_id.split("#")[2]
+    df_records["test"] = app_id.split("#")[0]
+    df_records["test_uuid"] = app_id.split("#")[1]
+    df_records["dataset"] = app_id.split("#")[2]
 
-    dfAll = pd.concat([dfAll, dfRecords], axis=0, ignore_index=True)
+    df_all = pd.concat([df_all, df_records], axis=0, ignore_index=True)
 
 print("Writing results to parquet file.")
-dfAll.to_parquet("results.parquet")
+df_all.to_parquet("results.parquet")
 print("Done!")

examples/evaluation/tru_langchain.py

@@ -8,7 +8,7 @@
 collection_name = "open_ai_512"
 
 vstore = tru_shared.get_astra_vector_store(framework, collection_name)
-chatModel = tru_shared.get_azure_chat_model(framework, "gpt-35-turbo", "0613")
+chat_model = tru_shared.get_azure_chat_model(framework, "gpt-35-turbo", "0613")
 embeddings = tru_shared.get_azure_embeddings_model(framework)
 
 prompt_template = """
@@ -22,7 +22,7 @@
 pipeline = (
     {"context": vstore.as_retriever(), "question": RunnablePassthrough()}
     | prompt
-    | chatModel
+    | chat_model
     | StrOutputParser()
 )
 

examples/evaluation/tru_llamaindex.py

@@ -13,10 +13,10 @@
 collection_name = "llama_512"
 
 vstore = tru_shared.get_astra_vector_store(framework, collection_name)
-chatModel = tru_shared.get_azure_chat_model(framework, "gpt-35-turbo", "0613")
+chat_model = tru_shared.get_azure_chat_model(framework, "gpt-35-turbo", "0613")
 embeddings = tru_shared.get_azure_embeddings_model(framework)
 
-service_context = ServiceContext.from_defaults(llm=chatModel, embed_model=embeddings)
+service_context = ServiceContext.from_defaults(llm=chat_model, embed_model=embeddings)
 storage_context = StorageContext.from_defaults(vector_store=vstore)
 vector_store_index = VectorStoreIndex.from_vector_store(
     vector_store=vstore, service_context=service_context

examples/evaluation/tru_shared.py

@@ -67,12 +67,12 @@ def init_tru():
 
 def get_feedback_functions(pipeline, golden_set):
     # Initialize provider class
-    azureOpenAI = AzureOpenAI(deployment_name="gpt-35-turbo")
+    azure_open_ai = AzureOpenAI(deployment_name="gpt-35-turbo")
 
     context = App.select_context(pipeline)
 
     # Define a groundedness feedback function
-    grounded = Groundedness(groundedness_provider=azureOpenAI)
+    grounded = Groundedness(groundedness_provider=azure_open_ai)
     f_groundedness = (
         Feedback(grounded.groundedness_measure_with_cot_reasons, name="groundedness")
         .on(context.collect())
@@ -82,19 +82,19 @@ def get_feedback_functions(pipeline, golden_set):
 
     # Question/answer relevance between overall question and answer.
     f_answer_relevance = Feedback(
-        azureOpenAI.relevance_with_cot_reasons, name="answer_relevance"
+        azure_open_ai.relevance_with_cot_reasons, name="answer_relevance"
     ).on_input_output()
 
     # Question/statement relevance between question and each context chunk.
     f_context_relevance = (
-        Feedback(azureOpenAI.qs_relevance_with_cot_reasons, name="context_relevance")
+        Feedback(azure_open_ai.qs_relevance_with_cot_reasons, name="context_relevance")
         .on_input()
         .on(context)
         .aggregate(np.mean)
     )
 
     # GroundTruth for comparing the Answer to the Ground-Truth Answer
-    ground_truth_collection = GroundTruthAgreement(golden_set, provider=azureOpenAI)
+    ground_truth_collection = GroundTruthAgreement(golden_set, provider=azure_open_ai)
     f_answer_correctness = Feedback(
         ground_truth_collection.agreement_measure, name="answer_correctness"
     ).on_input_output()
@@ -210,11 +210,11 @@ def execute_experiment(framework: Framework, pipeline, experiment_name: str):
 
     # use a short uuid to ensure that multiple experiments with the same name don't
     # collide in the DB
-    shortUuid = str(uuid.uuid4())[9:13]
+    short_uuid = str(uuid.uuid4())[9:13]
     datasets, golden_set = get_test_data()
 
     for dataset_name in datasets:
-        app_id = f"{experiment_name}#{shortUuid}#{dataset_name}"
+        app_id = f"{experiment_name}#{short_uuid}#{dataset_name}"
         tru_recorder = get_recorder(framework, pipeline, app_id, golden_set)
         for query in datasets[dataset_name]:
             try:

examples/notebooks/QA_with_cassio.ipynb

@@ -248,7 +248,7 @@
     "    if len(new_docs) > 0:\n",
     "        documents.extend(new_docs)\n",
     "\n",
-    "cassVStore = Cassandra.from_documents(\n",
+    "cass_vstore = Cassandra.from_documents(\n",
     "    documents=documents,\n",
     "    embedding=OpenAIEmbeddings(),\n",
     "    session=None,\n",
@@ -395,7 +395,7 @@
     "query = \"Who is Luchesi?\"\n",
     "\n",
     "# find matching documentation using similarity search\n",
-    "matched_docs = cassVStore.similarity_search(query=query, k=1)\n",
+    "matched_docs = cass_vstore.similarity_search(query=query, k=1)\n",
     "\n",
     "# print out the relevant context that an LLM will use to produce an answer\n",
     "for i, d in enumerate(matched_docs):\n",
@@ -450,7 +450,7 @@
     "from langchain.indexes.vectorstore import VectorStoreIndexWrapper\n",
     "\n",
     "llm = ChatOpenAI(model=\"gpt-3.5-turbo-1106\")\n",
-    "index = VectorStoreIndexWrapper(vectorstore=cassVStore)\n",
+    "index = VectorStoreIndexWrapper(vectorstore=cass_vstore)\n",
     "\n",
     "# Query the index for relevant vectors to our prompt\n",
     "query = \"Who is Luchesi?\"\n",
@@ -495,7 +495,7 @@
     "query = ChatPromptTemplate.from_template(query)\n",
     "\n",
     "qa = RetrievalQA.from_chain_type(\n",
-    "    llm=llm, retriever=cassVStore.as_retriever(), chain_type_kwargs={\"prompt\": query}\n",
+    "    llm=llm, retriever=cass_vstore.as_retriever(), chain_type_kwargs={\"prompt\": query}\n",
     ")\n",
     "\n",
     "result = qa.run(\"{question: Who is Luchesi?\")\n",
@@ -518,14 +518,14 @@
     "nbmake": {
      "post_cell_execute": [
       "# Deletes collection for test suite to allow each test to run with a fresh collection",
-      "cassVStore.delete_collection()"
+      "cass_vstore.delete_collection()"
      ]
     }
    },
    "outputs": [],
    "source": [
     "# WARNING: This will delete the collection and all documents in the collection\n",
-    "# cassVStore.delete_collection()"
+    "# cass_vstore.delete_collection()"
    ]
   }
  ],

libs/colbert/tests/unit_tests/test_colbert_baseline_embeddings.py

@@ -122,21 +122,21 @@ def test_colbert_embedding_against_vanilla_impl():
 
     embeddings_flat, _ = encoder.encode_passages(arctic_botany_chunks)
 
-    colbertSvc = ColbertEmbeddingModel(
+    colbert_svc = ColbertEmbeddingModel(
         checkpoint=DEFAULT_COLBERT_MODEL,
     )
-    embedded_chunks = colbertSvc.embed_texts(arctic_botany_chunks)
+    embedded_chunks = colbert_svc.embed_texts(arctic_botany_chunks)
 
     are_they_similar(embedded_chunks, embeddings_flat)
 
 
 def model_embedding(model: str):
     logging.info(f"test model compatibility {model}")
-    colbertSvc = ColbertEmbeddingModel(
+    colbert_svc = ColbertEmbeddingModel(
         checkpoint=model,
         query_maxlen=32,
     )
-    embeddings = colbertSvc.embed_texts(arctic_botany_chunks)
+    embeddings = colbert_svc.embed_texts(arctic_botany_chunks)
 
     assert len(embeddings) == 8
     n = 0
@@ -148,7 +148,7 @@ def model_embedding(model: str):
     assert n == 645
 
     # recall embeddings test
-    embedding = colbertSvc.embed_query(
+    embedding = colbert_svc.embed_query(
         query="What adaptations enable Arctic plants to survive and thrive "
         "in extremely cold temperatures and minimal sunlight?",
         query_maxlen=32,

libs/colbert/tests/unit_tests/test_colbert_embeddings.py

@@ -36,8 +36,8 @@ def test_colbert_query_embeddings():
     colbert = ColbertEmbeddingModel()
 
     embedding = colbert.embed_query("who is the president of the united states?")
-    queryTensor = torch.tensor(embedding)
-    assert queryTensor.shape == (12, 128)
+    query_tensor = torch.tensor(embedding)
+    assert query_tensor.shape == (12, 128)
 
     # test query encoding
     embedding = colbert.embed_query("test-query", query_maxlen=512)