Basic Breeze data structures
In this section, we will explore the basic data structures used to build Breeze programs. We will also discuss basic arithmetic operations that you can perform on these structures. The structures themselves are what you might expect from a numerical computing framework. If you have used MATLAB or Python's NumPy, you will instantly recognize them. The main ones you will probably be using are DenseVector
and DenseMatrix
. The sparse counterparts of these are SparseVector
and CSCMatrix
. These are optimized for use with large vectors and matrices that contain many zero elements. You can expect certain operations with these to be faster if your vectors and matrices fulfill the scarcity and largeness requirements. How sparse and large your data structures have to be for these to be worthwhile is not hard to guess. If you are dealing with sparse data structure, you should experiment and see if using SparseVector
or SparseMatrix
works faster.
DenseVector
Once in,...