Querying and Searching in ChromaDB

Welcome back! In the previous lesson, you learned how to insert and store embeddings in ChromaDB, setting the stage for more advanced operations. Today, we will focus on querying in ChromaDB, a crucial step in leveraging the power of vector-based search systems. Our goal is to guide you through performing a search query in ChromaDB and interpreting the results. This lesson will build on your existing knowledge and help you understand how to retrieve relevant information from your vector database efficiently.

Vector queries are at the heart of ChromaDB's search capabilities. Unlike traditional keyword searches, vector queries leverage the numerical representations of text, known as embeddings, to find semantically similar documents. In ChromaDB, a query is structured with query_texts, which are the input texts you want to search for, and n_results, which specifies the number of results you wish to retrieve. This approach allows you to perform more nuanced searches, capturing the meaning behind the text rather than just matching keywords.

Before proceeding to search in ChromaDB, let's define the data we'll be working with, as we learned in our previous lesson. Here is a code snippet that was deeply explained earlier:

# Sample documents
documents = [
    {"id": "doc1", "content": "ChromaDB is an open-source vector database."},
    {"id": "doc2", "content": "Vector databases help in efficient similarity search."},
    {"id": "doc3", "content": "Embedding models convert text into numerical representations."}
]

# Insert documents into ChromaDB
collection.add(
    documents=[doc["content"] for doc in documents],
    ids=[doc["id"] for doc in documents]
)

print("Inserted documents into ChromaDB.")

In this setup, we have a collection of sample documents, each with a unique identifier and content. These documents are inserted into ChromaDB, making them available for querying. This step ensures that our database is populated with data, allowing us to perform meaningful search operations in the subsequent example.

Let's walk through an example to demonstrate how to perform a search query in ChromaDB. Consider the following code snippet:

query_text = "What is ChromaDB?"

# Perform a search
results = collection.query(
    query_texts=[query_text],
    n_results=2
)

# Display results
for i, doc in enumerate(results["documents"][0]):
    print(f"Result {i+1}: {doc} (Score: {results['distances'][0][i]})")

In this example, we define a query_text with the question "What is ChromaDB?" and use the collection.query() method to perform the search. The query_texts parameter takes a list of input texts, and n_results specifies that we want the top two results. The results are then iterated over, displaying each document along with its distance score. The distance score indicates how closely the document matches the query, with lower scores representing more relevant results. When you run this code, you should see output similar to:

Result 1: ChromaDB is an open-source vector database. (Score: 0.123)
Result 2: Vector databases help in efficient similarity search. (Score: 0.456)