Previous: Test Functions, Up: Test and Demo Functions [Contents][Index]
Run example code block n associated with the function name.
If n is not specified, all examples are run.
The preferred location for example code blocks is embedded within the script
m-file immediately following the code that it exercises. Alternatively,
the examples may be stored in a file with the same name but no extension
located on Octave’s load path. To separate examples from regular script
code all lines are prefixed by %!
. Each example must also be
introduced by the keyword "demo"
flush left to the prefix with no
intervening spaces. The remainder of the example can contain arbitrary
Octave code. For example:
%!demo %! t = 0:0.01:2*pi; %! x = sin (t); %! plot (t, x); %! title ("one cycle of a sine wave"); %! #------------------------------------------------- %! # the figure window shows one cycle of a sine wave
Note that the code is displayed before it is executed so that a simple
comment at the end suffices for labeling what is being shown. For plots,
labeling can also be done with title
or text
. It is generally
not necessary to use disp
or printf
within the demo.
Demos are run in a stand-alone function environment with no access to external variables. This means that every demo must have separate initialization code. Alternatively, all demos can be combined into a single large demo with the code
%! input ("Press <enter> to continue: ", "s");
between the sections, but this usage is discouraged. Other techniques to
avoid multiple initialization blocks include using multiple plots with a new
figure
command between each plot, or using subplot
to put
multiple plots in the same window.
Finally, because demo
evaluates within a function context it is not
possible to define new functions within the code. Anonymous functions make
a good substitute in most instances. If function blocks must be
used then the code eval (example ("function", n))
will allow Octave
to see them. This has its own problems, however, as eval
only
evaluates one line or statement at a time. In this case the function
declaration must be wrapped with "if 1 <demo stuff> endif"
where
"if"
is on the same line as "demo"
. For example:
%!demo if 1 %! function y = f(x) %! y = x; %! endfunction %! f(3) %! endif
Display the code for example n associated with the function name, but do not run it.
If n is not specified, all examples are displayed.
When called with output arguments, the examples are returned in the form of a string s, with idx indicating the ending position of the various examples.
For a complete explanation see demo
.
Execute built-in tests for all m-files in the specified directory.
Test blocks in any C++ source files (*.cc) will also be executed for use with dynamically linked oct-file functions.
If no directory is specified, operate on all directories in Octave’s search path for functions.
Execute built-in demos for all m-files in the specified directory.
Demo blocks in any C++ source files (*.cc) will also be executed for use with dynamically linked oct-file functions.
If no directory is specified, operate on all directories in Octave’s search path for functions.
Determine the execution time of an expression (f) for various input values (n).
The n are log-spaced from 1 to max_n. For each n, an initialization expression (init) is computed to create any data needed for the test. If a second expression (f2) is given then the execution times of the two expressions are compared. When called without output arguments the results are printed to stdout and displayed graphically.
f
The code expression to evaluate.
max_n
The maximum test length to run. The default value is 100. Alternatively,
use [min_n, max_n]
or specify the n exactly with
[n1, n2, …, nk]
.
init
Initialization expression for function argument values. Use k for
the test number and n for the size of the test. This should compute
values for all variables used by f. Note that init will be
evaluated first for k = 0, so things which are constant throughout
the test series can be computed once. The default value is
x = randn (n, 1)
.
f2
An alternative expression to evaluate, so that the speed of two
expressions can be directly compared. The default is []
.
tol
Tolerance used to compare the results of expression f and expression
f2. If tol is positive, the tolerance is an absolute one.
If tol is negative, the tolerance is a relative one. The default is
eps
. If tol is Inf
, then no comparison will be made.
order
The time complexity of the expression O(a*n^p). This is a
structure with fields a
and p
.
n
The values n for which the expression was calculated AND the execution time was greater than zero.
T_f
The nonzero execution times recorded for the expression f in seconds.
T_f2
The nonzero execution times recorded for the expression f2 in seconds.
If required, the mean time ratio is simply mean (T_f ./ T_f2)
.
The slope of the execution time graph shows the approximate power of the asymptotic running time O(n^p). This power is plotted for the region over which it is approximated (the latter half of the graph). The estimated power is not very accurate, but should be sufficient to determine the general order of an algorithm. It should indicate if, for example, the implementation is unexpectedly O(n^2) rather than O(n) because it extends a vector each time through the loop rather than pre-allocating storage. In the current version of Octave, the following is not the expected O(n).
speed ("for i = 1:n, y{i} = x(i); endfor", "", [1000, 10000])
But it is if you preallocate the cell array y
:
speed ("for i = 1:n, y{i} = x(i); endfor", ... "x = rand (n, 1); y = cell (size (x));", [1000, 10000])
An attempt is made to approximate the cost of individual operations, but
it is wildly inaccurate. You can improve the stability somewhat by doing
more work for each n
. For example:
speed ("airy(x)", "x = rand (n, 10)", [10000, 100000])
When comparing two different expressions (f, f2), the slope of the line on the speedup ratio graph should be larger than 1 if the new expression is faster. Better algorithms have a shallow slope. Generally, vectorizing an algorithm will not change the slope of the execution time graph, but will shift it relative to the original. For example:
speed ("sum (x)", "", [10000, 100000], ... "v = 0; for i = 1:length (x), v += x(i); endfor")
The following is a more complex example. If there was an original version
of xcorr
using for loops and a second version using an FFT, then
one could compare the run speed for various lags as follows, or for a fixed
lag with varying vector lengths as follows:
speed ("xcorr (x, n)", "x = rand (128, 1);", 100, "xcorr_orig (x, n)", -100*eps) speed ("xcorr (x, 15)", "x = rand (20+n, 1);", 100, "xcorr_orig (x, n)", -100*eps)
Assuming one of the two versions is in xcorr_orig, this would compare their
speed and their output values. Note that the FFT version is not exact, so
one must specify an acceptable tolerance on the comparison 100*eps
.
In this case, the comparison should be computed relatively, as
abs ((x - y) ./ y)
rather than absolutely as
abs (x - y)
.
Type example ("speed") to see some real examples or demo ("speed") to run them.
Previous: Test Functions, Up: Test and Demo Functions [Contents][Index]