Function create_query_table

doctests/search_vss.py:260–299

Creates a query table.

(query, queries, encoded_queries, extra_params=None)

class="cm"># STEP_START define_bulk_query
def create_query_table(query, queries, encoded_queries, extra_params=None):
    class="st">"""
    Creates a query table.
    class="st">"""
    results_list = []
    for i, encoded_query in enumerate(encoded_queries):
        result_docs = (
            client.ft(class="st">"idx:bikes_vss")
            .search(
                query,
                {class="st">"query_vector": np.array(encoded_query, dtype=np.float32).tobytes()}
                | (extra_params if extra_params else {}),
            )
            .docs
        )
        for doc in result_docs:
            vector_score = round(1 - float(doc.vector_score), 2)
            results_list.append(
                {
                    class="st">"query": queries[i],
                    class="st">"score": vector_score,
                    class="st">"id": doc.id,
                    class="st">"brand": doc.brand,
                    class="st">"model": doc.model,
                    class="st">"description": doc.description,
                }
            )

    class="cm"># Optional: convert the table to Markdown using Pandas
    queries_table = pd.DataFrame(results_list)
    queries_table.sort_values(
        by=[class="st">"query", class="st">"score"], ascending=[True, False], inplace=True
    )
    queries_table[class="st">"query"] = queries_table.groupby(class="st">"query")[class="st">"query"].transform(
        lambda x: [x.iloc[0]] + [class="st">""] * (len(x) - 1)
    )
    queries_table[class="st">"description"] = queries_table[class="st">"description"].apply(
        lambda x: (x[:497] + class="st">"...") if len(x) > 500 else x
    )
    return queries_table.to_markdown(index=False)


class="cm"># STEP_END