C++11: should I use valarray or vector for numerical computing

Question

The question of vector vs valarray has already been asked here. My question refers specifically to the case of C++11. I have been reading a "A Tour of C++" and "The C++ Programming Language". Both books are written by Bjarne Stroustrup. In the first book the author explains that std::valarray should be preferred for numerical computing (Chapter 12). But then in the second book, in chapter 29, the author implements a Matrix class in terms of a std::vector.

Also by doing a bit of googling, it seems that performance-wise, a std::vector is just as fast as dynamically allocated "raw arrays".

So in the context of C++11, which container should be preferred for numerical computing?

My take on this would be that since std::vector provides fast access to its contents using the operator[] (which returns a reference to the data with no bounds checking) and that it is also safer to use a std::vector over a dynamically allocated array, std::vector should be preferred. Also, from C++11 onwards:

• std::vector provides direct access to the underlying data using std::vector::data()
• std::vector can be resized in order to use less memory using std::vector::shrink_to_fit()

Answer 1

valarray has the nice functionality, that you easily can apply mathematical functions element-wise and you have better slicing abilities. You can e.g. do v3 = sin(v2 + v1*3)

Nevertheless, if you really want to do scientific computing, consider using a library such as Eigen