tfidf retriever (#5114)

Co-authored-by: vempaliakhil96 <vempaliakhil96@gmail.com>

docs/modules/indexes/retrievers/examples/tf_idf.ipynb

@@ -16,17 +16,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "id": "a801b57c",
    "metadata": {},
    "outputs": [],
    "source": [
-    "# !pip install scikit-learn"
+    "# !pip install scikit-learn\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "id": "393ac030",
    "metadata": {
     "tags": []
@@ -46,7 +46,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 3,
    "id": "98b1c017",
    "metadata": {
     "tags": []
@@ -56,6 +56,27 @@
     "retriever = TFIDFRetriever.from_texts([\"foo\", \"bar\", \"world\", \"hello\", \"foo bar\"])"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "c016b266",
+   "metadata": {},
+   "source": [
+    "## Create a New Retriever with Documents\n",
+    "\n",
+    "You can now create a new retriever with the documents you created."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "53af4f00",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from langchain.schema import Document\n",
+    "retriever = TFIDFRetriever.from_documents([Document(page_content=\"foo\"), Document(page_content=\"bar\"), Document(page_content=\"world\"), Document(page_content=\"hello\"), Document(page_content=\"foo bar\")])"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "08437fa2",
@@ -68,7 +89,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 6,
    "id": "c0455218",
    "metadata": {
     "tags": []
@@ -80,7 +101,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 7,
    "id": "7dfa5c29",
    "metadata": {
     "tags": []
@@ -95,7 +116,7 @@
        " Document(page_content='world', metadata={})]"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 7,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -103,14 +124,6 @@
    "source": [
     "result"
    ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "74bd9256",
-   "metadata": {},
-   "outputs": [],
-   "source": []
   }
  ],
  "metadata": {
@@ -129,7 +142,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.6"
+   "version": "3.11.3"
   }
  },
  "nbformat": 4,

langchain/retrievers/tfidf.py

@@ -2,7 +2,9 @@
 
 Largely based on
 https://github.com/asvskartheek/Text-Retrieval/blob/master/TF-IDF%20Search%20Engine%20(SKLEARN).ipynb"""
-from typing import Any, Dict, List, Optional
+from __future__ import annotations
+
+from typing import Any, Dict, Iterable, List, Optional
 
 from pydantic import BaseModel
 
@@ -23,18 +25,39 @@ class TFIDFRetriever(BaseRetriever, BaseModel):
     @classmethod
     def from_texts(
         cls,
-        texts: List[str],
+        texts: Iterable[str],
+        metadatas: Optional[Iterable[dict]] = None,
         tfidf_params: Optional[Dict[str, Any]] = None,
-        **kwargs: Any
-    ) -> "TFIDFRetriever":
-        from sklearn.feature_extraction.text import TfidfVectorizer
+        **kwargs: Any,
+    ) -> TFIDFRetriever:
+        try:
+            from sklearn.feature_extraction.text import TfidfVectorizer
+        except ImportError:
+            raise ImportError(
+                "Could not import scikit-learn, please install with `pip install "
+                "scikit-learn`."
+            )
 
         tfidf_params = tfidf_params or {}
         vectorizer = TfidfVectorizer(**tfidf_params)
         tfidf_array = vectorizer.fit_transform(texts)
-        docs = [Document(page_content=t) for t in texts]
+        metadatas = metadatas or ({} for _ in texts)
+        docs = [Document(page_content=t, metadata=m) for t, m in zip(texts, metadatas)]
         return cls(vectorizer=vectorizer, docs=docs, tfidf_array=tfidf_array, **kwargs)
 
+    @classmethod
+    def from_documents(
+        cls,
+        documents: Iterable[Document],
+        *,
+        tfidf_params: Optional[Dict[str, Any]] = None,
+        **kwargs: Any,
+    ) -> TFIDFRetriever:
+        texts, metadatas = zip(*((d.page_content, d.metadata) for d in documents))
+        return cls.from_texts(
+            texts=texts, tfidf_params=tfidf_params, metadatas=metadatas, **kwargs
+        )
+
     def get_relevant_documents(self, query: str) -> List[Document]:
         from sklearn.metrics.pairwise import cosine_similarity
 

poetry.lock

@@ -6874,7 +6874,6 @@ files = [
     {file = "pylance-0.4.12-cp38-abi3-macosx_10_15_x86_64.whl", hash = "sha256:2b86fb8dccc03094c0db37bef0d91bda60e8eb0d1eddf245c6971450c8d8a53f"},
     {file = "pylance-0.4.12-cp38-abi3-macosx_11_0_arm64.whl", hash = "sha256:0bc82914b13204187d673b5f3d45f93219c38a0e9d0542ba251074f639669789"},
     {file = "pylance-0.4.12-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:5a4bcce77f99ecd4cbebbadb01e58d5d8138d40eb56bdcdbc3b20b0475e7a472"},
-    {file = "pylance-0.4.12-cp38-abi3-win_amd64.whl", hash = "sha256:9616931c5300030adb9626d22515710a127d1e46a46737a7a0f980b52f13627c"},
 ]
 
 [package.dependencies]
@@ -10857,7 +10856,7 @@ azure = ["azure-ai-formrecognizer", "azure-ai-vision", "azure-cognitiveservices-
 cohere = ["cohere"]
 docarray = ["docarray"]
 embeddings = ["sentence-transformers"]
-extended-testing = ["atlassian-python-api", "beautifulsoup4", "beautifulsoup4", "chardet", "gql", "html2text", "jq", "lxml", "pandas", "pdfminer-six", "psychicapi", "pymupdf", "pypdf", "pypdfium2", "requests-toolbelt", "telethon", "tqdm", "zep-python"]
+extended-testing = ["atlassian-python-api", "beautifulsoup4", "beautifulsoup4", "chardet", "gql", "html2text", "jq", "lxml", "pandas", "pdfminer-six", "psychicapi", "pymupdf", "pypdf", "pypdfium2", "requests-toolbelt", "scikit-learn", "telethon", "tqdm", "zep-python"]
 llms = ["anthropic", "cohere", "huggingface_hub", "manifest-ml", "nlpcloud", "openai", "openlm", "torch", "transformers"]
 openai = ["openai", "tiktoken"]
 qdrant = ["qdrant-client"]
@@ -10866,4 +10865,4 @@ text-helpers = ["chardet"]
 [metadata]
 lock-version = "2.0"
 python-versions = ">=3.8.1,<4.0"
-content-hash = "196588e10bb33939f5bae294a194ad01e803f40ed1087fe6a7a4b87e8d80712b"
+content-hash = "640f7e8102328d7ec3f56778d7cdb76b4846fc407c99606e0aec31833bc3933e"

pyproject.toml

@@ -93,6 +93,7 @@ langkit = {version = ">=0.0.1.dev3, <0.1.0", optional = true}
 chardet = {version="^5.1.0", optional=true}
 requests-toolbelt = {version = "^1.0.0", optional = true}
 openlm = {version = "^0.0.5", optional = true}
+scikit-learn = {version = "^1.2.2", optional = true}
 azure-ai-formrecognizer = {version = "^3.2.1", optional = true}
 azure-ai-vision = {version = "^0.11.1b1", optional = true}
 azure-cognitiveservices-speech = {version = "^1.28.0", optional = true}
@@ -274,7 +275,8 @@ extended_testing = [
  "zep-python",
  "gql",
  "requests_toolbelt",
- "html2text"
+ "html2text",
+ "scikit-learn",
 ]
 
 [tool.ruff]

tests/unit_tests/retrievers/test_tfidf.py

@@ -1,6 +1,10 @@
+import pytest
+
 from langchain.retrievers.tfidf import TFIDFRetriever
+from langchain.schema import Document
 
 
+@pytest.mark.requires("sklearn")
 def test_from_texts() -> None:
     input_texts = ["I have a pen.", "Do you have a pen?", "I have a bag."]
     tfidf_retriever = TFIDFRetriever.from_texts(texts=input_texts)
@@ -8,6 +12,7 @@ def test_from_texts() -> None:
     assert tfidf_retriever.tfidf_array.toarray().shape == (3, 5)
 
 
+@pytest.mark.requires("sklearn")
 def test_from_texts_with_tfidf_params() -> None:
     input_texts = ["I have a pen.", "Do you have a pen?", "I have a bag."]
     tfidf_retriever = TFIDFRetriever.from_texts(
@@ -15,3 +20,15 @@ def test_from_texts_with_tfidf_params() -> None:
     )
     # should count only multiple words (have, pan)
     assert tfidf_retriever.tfidf_array.toarray().shape == (3, 2)
+
+
+@pytest.mark.requires("sklearn")
+def test_from_documents() -> None:
+    input_docs = [
+        Document(page_content="I have a pen."),
+        Document(page_content="Do you have a pen?"),
+        Document(page_content="I have a bag."),
+    ]
+    tfidf_retriever = TFIDFRetriever.from_documents(documents=input_docs)
+    assert len(tfidf_retriever.docs) == 3
+    assert tfidf_retriever.tfidf_array.toarray().shape == (3, 5)