Mercurial > hg > aim92
comparison man/man1/scale.1 @ 0:5242703e91d3 tip
Initial checkin for AIM92 aimR8.2 (last updated May 1997).
author | tomwalters |
---|---|
date | Fri, 20 May 2011 15:19:45 +0100 |
parents | |
children |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:5242703e91d3 |
---|---|
1 .TH SCALE 1 "1 September 1993" | |
2 | |
3 .SH NAME | |
4 scale \- Shift and scale a binary stream by given constants or to | |
5 fit a given range (top - bottom). | |
6 | |
7 .SH SYNTAX | |
8 scale [options] [file] | |
9 | |
10 scale -h -for help with options and their defaults. | |
11 | |
12 .SH DESCRIPTION | |
13 Shifting and scaling is done for all input x using the formula: | |
14 .nf | |
15 y = ( x - shift ) * scale | |
16 .fi | |
17 | |
18 .SH OPTIONS | |
19 | |
20 1. top, bottom. | |
21 | |
22 If neither shift nor scale factor are given (ie. both options are "off"), | |
23 then values are | |
24 chosen so that the output fits the range specified by `top' and `bottom'. | |
25 For an input file which ranges between extreme values of `max' and `min', the | |
26 shift and scale factors to give a required range between `top' and `bottom' | |
27 are found by simultaneously solving: | |
28 .nf | |
29 top = ( max - shift ) * scale | |
30 bottom = ( min - shift ) * scale | |
31 .fi | |
32 | |
33 If top < bottom then the input is inverted in the resulting range. | |
34 | |
35 2. normalize. | |
36 | |
37 Values of shift and scale are chosen so that the data is normalized for | |
38 zero mean and unit standard deviation. | |
39 | |
40 .nf | |
41 shift = mean | |
42 scale = 1 / sqrt( variance ) | |
43 .fi | |
44 | |
45 3. shift, scale. | |
46 | |
47 In either or both shift and scale are given (ie. are not "off") then | |
48 the formula is applied directly irrespective of any given top or bottom | |
49 parameters. | |
50 If only a shift term is given, then the default scale becomes 1.0. | |
51 If only a scale factor is given, then the default shift becomes 0. | |
52 | |
53 4. type. | |
54 | |
55 The input and output data type are selected using "type=<type>", where | |
56 <type> = { char, short, int, float, double, ASCII }. | |
57 The ASCII type is interpreted as one number per line. | |
58 | |
59 5. output. | |
60 | |
61 If no filename arguments are given to the program then one input file is | |
62 expected on the stdin, otherwise each given filename is input and scaled in | |
63 turn. | |
64 If "output=off" then program output overwrites each | |
65 respective input file. (The respective output for the stdin is the stdout). | |
66 If the `output' option is given a value (eg. output=filename) then this is | |
67 used as a filename and all scaled input files are written to this output file. | |
68 (The default output filename is the stdout). | |
69 Special value "output=print" suppresses output and prints the shift and | |
70 scale parameters on the stdout. | |
71 | |
72 6. range. | |
73 | |
74 The range option specifies the stretch of the wave for scaling and output by: | |
75 | |
76 range=a[-b] | |
77 | |
78 The upper limit `b' is optional, and when it is missing then the range | |
79 is a single sample, otherwise `a' and `b' are inclusive range limits. | |
80 The strings "min" and "max" are recognised as extreme limits ("min" = | |
81 start of file, "max" = end of file), otherwise the values of `a' and `b' | |
82 are sample numbers: 0,1,2,... Values given with time units (ms or s) | |
83 are converted to samples using the given samplerate option. | |
84 | |
85 .nf | |
86 range=min-10ms -the first 10ms of each input file. | |
87 range=128-max -from the 128'th sample to the end of each file. | |
88 range=8ms-16ms -between 8ms and 16ms inclusive for each file. | |
89 .fi | |
90 | |
91 .SH "SEE ALSO" | |
92 options op merge | |
93 | |
94 .SH COPYRIGHT | |
95 .LP | |
96 Copyright (c) Applied Psychology Unit, Medical Research Council, 1995 | |
97 .LP | |
98 Permission to use, copy, modify, and distribute this software without fee | |
99 is hereby granted for research purposes, provided that this copyright | |
100 notice appears in all copies and in all supporting documentation, and that | |
101 the software is not redistributed for any fee (except for a nominal | |
102 shipping charge). Anyone wanting to incorporate all or part of this | |
103 software in a commercial product must obtain a license from the Medical | |
104 Research Council. | |
105 .LP | |
106 The MRC makes no representations about the suitability of this | |
107 software for any purpose. It is provided "as is" without express or | |
108 implied warranty. | |
109 .LP | |
110 THE MRC DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING | |
111 ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL | |
112 THE A.P.U. BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES | |
113 OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, | |
114 WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, | |
115 ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS | |
116 SOFTWARE. | |
117 |