Numerical Recipes is a series of books [1] [2] [3][4][5][6].
The series describes the
and some other useful algorithms.
The first books of the series were published in the past century, and the algorithms did not update much since that time. Although the algorithms were breakthrough when they just appear, various authors mention, that the performance is significantly lower than that of modern scientific packages like GNU Scientific Library, LAPACK and others [7].
An important disadvantage is non-free distribution. It can be useful to understand, how does an algorithm works; then in order to make some software and distribute it, the algorithm should be re-written. The book could be recommended for educational purposes, to learn the principles of scientific computation, rather than for high performance scientific simulation.
The advantage of the recipes is that they are self-consistent; usually, it is sufficient to load a program with very few dependencies; such a loading does not require any installation or adjusting parameters.