Contextual compression in LangChain - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ component_behavior

intermediate

What is the output of this LangChain Contextual Compression Retriever code?

Given the following code snippet using LangChain's ContextualCompressionRetriever, what will be the output when retrieving documents for the query 'Explain AI'?

LangChain

from langchain.document_loaders import TextLoader
from langchain.indexes import VectorstoreIndexCreator
from langchain.retrievers import ContextualCompressionRetriever
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import OpenAIEmbeddings
from langchain.vectorstores import FAISS

# Load documents
loader = TextLoader('ai_text.txt')
docs = loader.load()

# Split documents
splitter = RecursiveCharacterTextSplitter(chunk_size=100, chunk_overlap=10)
split_docs = splitter.split_documents(docs)

# Create vectorstore
embeddings = OpenAIEmbeddings()
vectorstore = FAISS.from_documents(split_docs, embeddings)

# Create compression retriever
compression_retriever = ContextualCompressionRetriever(
    base_compressor=None,  # No compressor for simplicity
    base_retriever=vectorstore.as_retriever()
)

# Retrieve compressed docs
results = compression_retriever.get_relevant_documents('Explain AI')
print(len(results))

A Raises a TypeError because base_compressor cannot be None

B Prints the number of documents retrieved, equal to the number of chunks matching 'Explain AI' query

C Prints zero because no compressor filters out all documents

D Raises a NameError because ContextualCompressionRetriever is not imported

💡 Hint

Think about how the retriever works when no compressor is provided.

When base_compressor is None, ContextualCompressionRetriever behaves like the base retriever, returning all matching documents. So the output is the count of matching chunks.

📝 Syntax

intermediate

Which option correctly initializes a ContextualCompressionRetriever with a compressor?

Select the code snippet that correctly creates a ContextualCompressionRetriever with a base compressor and base retriever.

A compression_retriever = ContextualCompressionRetriever(compressor=compressor, retriever=retriever)

B compression_retriever = ContextualCompressionRetriever(base_compressor=compressor)

C compression_retriever = ContextualCompressionRetriever(base_compressor=compressor, base_retriever=retriever)

D compression_retriever = ContextualCompressionRetriever(base_retriever=retriever)

💡 Hint

Check the exact parameter names required by ContextualCompressionRetriever.

ContextualCompressionRetriever requires both base_compressor and base_retriever parameters. Option C uses the correct parameter names and provides both.

🔧 Debug

advanced

Why does this ContextualCompressionRetriever code raise a ValueError?

Consider this code snippet: from langchain.retrievers import ContextualCompressionRetriever retriever = ContextualCompressionRetriever(base_compressor=compressor) What is the cause of the ValueError?

A Missing base_retriever argument causes ValueError because it is required

B compressor variable is undefined causing ValueError

C ContextualCompressionRetriever cannot be instantiated directly causing ValueError

D ValueError occurs because base_compressor must be None

💡 Hint

Check the required parameters for ContextualCompressionRetriever initialization.

ContextualCompressionRetriever requires both base_compressor and base_retriever. Omitting base_retriever causes a ValueError.

❓ state_output

advanced

What is the effect of using a compressor in ContextualCompressionRetriever on retrieved documents?

If a compressor is used in ContextualCompressionRetriever, how does it affect the documents returned for a query?

A The compressor duplicates documents, increasing the number of results

B The compressor has no effect; all documents are returned unchanged

C The compressor filters out all documents, returning an empty list

D The compressor reduces document size by summarizing, so fewer but more relevant documents are returned

💡 Hint

Think about what compression means in this context.

The compressor summarizes or compresses documents before retrieval, so fewer but more focused documents are returned.

🧠 Conceptual

expert

Which statement best describes the role of ContextualCompressionRetriever in LangChain?

Choose the statement that best explains what ContextualCompressionRetriever does in LangChain.

A It wraps a retriever and a compressor to return compressed, relevant documents for a query

B It replaces the retriever and compressor with a single unified model

C It retrieves documents but does not support compression

D It only compresses documents without retrieving them

💡 Hint

Consider how retrieval and compression work together.

ContextualCompressionRetriever combines a base retriever and a compressor to return compressed relevant documents, improving efficiency.