blob: 4d36851f84393b3c4e56253674f63a0f24e47ddb (
plain) (
blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
|
.help raverage May07 lists
.ih
NAME
raverage -- running average, standard deviation, and envelope
.ih
USAGE
raverage input nwin
.ih
PARAMETERS
.ls input
Input one or two column list of numbers. Any line that can't be read
as one or two numbers is ignored which means comments are allowed. The
special name "STDIN" may be used to read the numbers from the standard
input pipe or redirection.
.le
.ls nwin
The number of values in the running average window.
.le
.ls sort = no
Numerically sort the first column of the input list by increasing value?
This is done in an temporary file and the
actual input file is not modified.
.le
.ls nsig = 0
The number of standard deviations below and above the average for the
envelope columns. If the value is greater than zero two extra columns
are output formed by subtracting and adding this number of standard
deviations to the average value. If the value is zero then the
columns not be written.
.le
.ls fd1, fd2
Internal parameters.
.le
.ih
DESCRIPTION
This task computes the running average and standard deviation of a
one or two column list. For a one column list the ordinal value is
added. Note that the ordinal is only for the lines that are successfully
read so any comments are not counted. So the internal list is always
two columns.
The input may be a physical file or the standard input. The standard
input is specified by the special name "STDIN". All the input values
are read and stored in a temporary file prior to computing the output.
A temporary file is also used if the input is to be numerically sorted
by increasing value of the first column. Note that the sorting is done
before adding the implied ordinal for one column lists.
The output has four or six columns depending on whether \fInsig\fR is
zero or greater than zero.
.nf
average1 average1 stddev number [lower upper]
average1 - the running average of the first column
average2 - the running average of the second column
stddev - standard deviation of the second column
number - number of values in the statistic
lower - optional lower envelope value
upper - optional upper envelope value
.fi
The "number" of values may be less than the window if the window size is
larger than the list.
The number of lines will generally be less than the input because there is
no boundary extension. In other words the first output value is computed
after the first \fInwin\fR values have been read and the last output value
is computed when the end of the list is reached.
The envelope columns are computed when \fInsig\fR is greater than zero.
The values are
.nf
lower = average2 - nsig * stddev
upper = average2 + nsig * stddev
.fi
In many cases the data is intended to represent a scatter plot and one
wants to show the trend and envelope as a function of the first column.
This is where the sorting and envelope options are useful.
.ih
EXAMPLES
1. Compute the running average with a window of 100 values on the list of
numbers in file "numbers".
.nf
cl> raverage numbers 100
.fi
2. Do this using the standard input. In this example use random numbers.
.nf
cl> urand 100 1 | raverage STDIN 90
.fi
3. Make a scatter plot of a two column list with the trend and envelope
overplotted.
.nf
cl> fields numbers 1,3 | graph point+
cl> fields numbers 1,3 | raverage STDIN 100 sort+ nsig=3 > tmp
cl> fields tmp 1,2 | graph append+
cl> fields tmp 1,5 | graph append+
cl> fields tmp 1,6 | graph append+
.fi
.ih
SEE ALSO
average, boxcar
.endhelp
|
