StateVectorDB

StateVector is a highly efficient vector database leveraging classical and quantum algorithms to optimize cosine similarity calculations. It supports various data types such as text, images, and audio, providing fast and accurate similarity searches.

Features

• Vector Representation: Converts text, images, and audio into vector representations using algorithms like Word2Vec.OpenCV，Librosa
• Similarity Calculation: Computes cosine similarity using both classical and quantum-optimized algorithms.
• Indexing: Builds efficient indexes using QuantumKdTreeSearch Trees for quick retrieval of similar vectors.
• Search: Performs nearest neighbor search for similarity queries.

Installation

1. Clone the repository:

   git clone https://github.com/crystal-tensor/StateVector.git
   cd StateVector

2. Install the required dependencies:

   pip install -r requirements.txt

3. You can install StateVector directly from GitHub:

pip install git+https://github.com/crystal-tensor/StateVector.git

or

pip install statevector

Usage

Text Vectorization

from vector_representation.text_vectorizer import TextVectorizer

vectorizer = TextVectorizer('model/word2vec_model.bin')
vector = vectorizer.vectorize("example text")

Cosine Similarity

from similarity.cosine_similarity import CosineSimilarity

similarity = CosineSimilarity.compute(vector1, vector2)

Quantum Cosine Similarity

from similarity.quantum_cosine_similarity import QuantumCosineSimilarity

quantum_similarity = QuantumCosineSimilarity()
similarity = quantum_similarity.compute(vector1, vector2)

KD Tree Indexing

from index.kd_tree import KDTreeIndex

index = KDTreeIndex(data)
distance, index = index.query(vector, k=5)

Nearest Neighbor Search

from search.nearest_neighbor import NearestNeighborSearch

search = NearestNeighborSearch(index)
results = search.search(vector, k=5)

Documentation

See the docs directory for detailed documentation on each module.

---

StateVectorDB

StateVector 是一个高效的向量数据库，利用经典和量子算法优化余弦相似性计算。它支持文本、图像和音频等多种数据类型，提供快速准确的相似性搜索。

特性

• 向量表示：使用 Word2Vec，OpenCV，Librosa等算法库将文本、图像和音频转换为向量表示。
• 相似度计算：使用经典和量子优化算法计算余弦相似度。
• 索引：使用 HyperLogKD 树构建高效索引，快速检索相似向量。
• 搜索：执行最近邻搜索以进行相似性查询。

安装

1. 克隆仓库：

   git clone https://github.com/crystal-tensor/StateVector.git
   cd StateVector

2. 安装所需依赖：

   pip install -r requirements.txt

3、安装向量数据库

pip install statevector

使用方法

文本向量化

from vector_representation.text_vectorizer import TextVectorizer

vectorizer = TextVectorizer('model/word2vec_model.bin')
vector = vectorizer.vectorize("example text")

余弦相似度

from similarity.cosine_similarity import CosineSimilarity

similarity = CosineSimilarity.compute(vector1, vector2)

量子余弦相似度

from similarity.quantum_cosine_similarity import QuantumCosineSimilarity

quantum_similarity = QuantumCosineSimilarity()
similarity = quantum_similarity.compute(vector1, vector2)

KD 树索引

from index.kd_tree import KDTreeIndex

index = KDTreeIndex(data)
distance, index = index.query(vector, k=5)

最近邻搜索

from search.nearest_neighbor import NearestNeighborSearch

search = NearestNeighborSearch(index)
results = search.search(vector, k=5)

文档

详见 docs 目录下各模块的详细文档。