SPECTRAL METHODS IN MATLAB - ERRATA (Some of these have been corrected in later printing, so not all may appear in a particular copy of the book.) p 26, p6.m: in the "axis" command, 0 5 should be -0.1 5 p 34, l 4: ||u(x)-p(x)|| should be ||u-p|| p 46, (5.9) z - sqrt... should be z + sqrt... p 48, Thm 6: \phi_f should be \phi_u (3 times) p 66, Program 15 and Output 15: 4/\pi^2 should be \pi^2/4 p 70, Program 16 and nearby discussion: x and y have been interchanged p 72, displayed equation: 1/2 should be \pi/2 p 78, special formulas for endpoints: \hat v should be \hat V p 79, top: the switch from p( ) to q( ) is unnecessary p 85, Ex. 8.1: exponent 1/N should be 1/n. p 99, Exercise 9.5: the boundary conditions in this problem are wrong p 108, l 5: <= should be >= (or the signs should be switched) p 109, Program 26 and Output 26: 2/\pi/N should be 2N/\pi p 109, after (10.4): "inverse" should be "reciprocal" p 109, eq (10.4): a^3*t should be a^3*t/2, and 2a^2 should be a^2 twice in paragraph following p 119, between displayed eqs: N_{\theta/2} should be {N_\theta}/2 (twice) chap. 11: The reversed indexing used to define D2 and E2 is correct in Program 28, but this is not described in the text. p 132, line above (12.9): d\theta is missing twice from equation involving dt p 137, (13.4): u(-1,t) = -1, not 0 p 139, Program 34: in first comment line, u_xx should be eps*u_xx p 139, Program 34: (v-x) can be replaced by v in line marked % Euler p 140, Exercise 13.1: it should be specified that the problem is linear p 143, Program 37: the definition of dt needs to be moved to before where it is used (this has been fixed in the program online) p 144, Exercise 13.2: Output/Program 32 should be 34 p 155, [BaBe00]: this label should be [BaBe99] I am grateful to Prof. Jean-Paul Berrut of the University of Fribourg, Switzerland for finding many of these errors. If you find further errors in the book, please let me know: trefethen@maths.ox.ac.uk - Nick Trefethen