In octave, what are the error tolerance options in numerical integration?

Question

In Octave, you can call quad_options to get:

quad_options()

Options for Quad include:

keyword                                             value
-------                                             -----
absolute tolerance                                  1.49012e-08
relative tolerance                                  1.49012e-08
single precision absolute tolerance                 0.000345267
single precision relative tolerance                 0.000345267

Can someone explain the difference between these four kinds of error tolerances?

Answer 1

For the difference between absolute and relative tolerance see the answers to this question. The last two are the same but when you are using data with single precision instead of double.