1
%++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2
%.COPYRIGHT (c) 1990 European Southern Observatory
4
%.AUTHOR MR,FM,KB, IPG/ESO
5
%.KEYWORDS MIDAS, help files, CREATE/RANDOM_IMAGE
6
%.PURPOSE On-line help file for the command: CREATE/RANDOM_IMAGE
7
%.VERSION 1.0 03-JAN-1986 : Creation, MR,FM,KB
8
%----------------------------------------------------------------
12
CREATE/RANDOM 14-JUL-2003 KB
14
CREATE/RANDOM name [dims] [starts,steps] [func_type] [coefs] [seed]
15
CREATE/RANDOM name = ref_frame [func_type] [coefs] [seed]
19
Create a new image (= bulk data frame + standard descriptors)
20
with pixel values drawn from a random distribution.
23
CREATE/RANDOM name [dims] [starts,steps] [func_type] [coefs] [seed]
24
CREATE/RANDOM name = ref_frame [func_type] [coefs] [seed]
26
name = name of new frame
28
ref_frame = reference frame,
29
take descriptors NAXIS, START, STEP, ... from that frame
31
dims = NAXIS, NPIX(1),...,NPIX(NAXIS) (NAXIS max. 3);
34
starts,steps = START(1),...,START(NAXIS),STEP(1),...,STEP(NAXIS)
35
defaulted to: 0. for START(i), 1. for STEP(i)
37
func_type = type,sub_type
38
of distribution function used for image creation
39
type: U(NIFORM) for uniform distribution
40
G(AUSS) for Gaussian d.
41
E(XPONENTIAL) for exponential d.
42
B(INOMIAL) for binomial d.
43
P(OISSON) for Poisson d.
44
C(AUCHY) for Cauchy d.
46
F(ibonacci) for lagged Fibonacci random no. generator
47
S(tandard) for minimal standard uniformF random no. g.
48
defaulted to UNIFORM,Fibonacci
50
coefs = coefficients for the distributions above
51
number and meaning depends on the parameter func_type;
54
seed = Integer number used as root of the pseudo-random
55
series. Defaulted to the current time in seconds/1237,
56
as provided by M$SECS().
62
The random number generators have been reimplemented and do not
63
use the NAG library anymore.
65
For `funct_type' GAUSS the coefficients represent m,s (mean and
66
sigma), for UNIFORM the coefficients represent a,b the endpoints
67
of the interval [a,b].
68
For BINOMIAL the coefficients represent the no. of trials and
69
the probabiltiy of an event.
70
For EXPONENTIAL, POISSON, CAUCHY the first coefficient
71
represents m the mean (2nd coeff. is ignored).
73
The sub_type refers to the type of r.n.g. used for the generation
74
of the uniformly distributed values. Since also the other
75
distributions are based implicitely on uniform random numbers,
76
the sub_type can be applied to all distributions.
80
CREATE/RANDOM new 2,100,100 -200.,-180.,22.,22. uniform,standard
81
Will create the frame `new.bdf' with 100*100 pixels,
82
start values -200.,-180., stepsize 22.,22.
83
the pixel values of `new.bdf' will be determined according
84
to a uniform distribution (using minimal standard algorithm)
85
over the interval [0,1].
87
CREATE/RANDOM new = sombrero GAUSS 12.5,1.25
88
Will create the image `new.bdf' with same size and standard
89
descriptors as the frame `sombrero.bdf'.
90
The pixel values of `new.bdf' will be determined according
91
to a Gaussian distribution with mean = 12.5 and
92
standard deviation = 1.25.
93
The uniformly distributed random values used for the calculation
94
of the Gaussian distribution are generated via the Fibonacci
95
algorithm (the default).
97
CREATE/RANDOM bino ? ? binomial,st 12,0.66 12347
98
Will create the image `bino.bdf' with size 64 x 64,
99
start = 0.0, 0.0 and step = 1.0, 1.0 according to a binomial
100
distribution using 12 trials and probability = 0.66.
101
The underlying uniform r.n.g. uses the minimal standard algorithm
102
and seed value = 12347.