EOS1实现3种版本的向量内积） ,返回double精度数值） ,比较使用不同大下和密度的向量来对比相关的速度） ,-In large-scale optimization problems, different vectors and matrices are typically sparse, i.e. only very few elements are non-zero. Typically vectors with thousands of elements may contain o