Gnuplot » Historique » Version 4
François Rioult, 03/02/2011 20:15
1 | 1 | François Rioult | h1. Gnuplot |
---|---|---|---|
2 | |||
3 | h2. Synopsis |
||
4 | |||
5 | This operator has no parameter, two inputs (the GNUplot script and the data) and one output (the plot). It launches GNUplot for generating a .PNG image. |
||
6 | |||
7 | h2. Parameters |
||
8 | |||
9 | None. |
||
10 | |||
11 | h2. Inputs |
||
12 | |||
13 | 2 | François Rioult | 2 input files : |
14 | 4 | François Rioult | * the first is the @GNUplot@ script. This script uses a @@file@@ string, that is dynamically replaced with the real name by the script. Example |
15 | <pre>set title "complexite de l'extraction des frequents" |
||
16 | set logscale |
||
17 | set style data lines |
||
18 | set xlabel 'support minimum' |
||
19 | set ylabel 'nombre de frequents' |
||
20 | set y2tics |
||
21 | set y2label 'temps extraction en s.' |
||
22 | plot '@file@' using 1:2 title 'nombre de frequents' axis x1y1, '@file@' using 1:3 title 'temps extraction' axis x1y2, '@file@' using 1:($2/$3) title 'rapport temps/frequent' |
||
23 | </pre> |
||
24 | 2 | François Rioult | * the second is the data to be plot. |
25 | 1 | François Rioult | |
26 | h2. Outputs |
||
27 | |||
28 | 1 output: the .PNG image representing the plot. |
||
29 | |||
30 | h2. Shell Code |
||
31 | |||
32 | <pre> |
||
33 | #!/bin/bash |
||
34 | |||
35 | # this shell uses gnuplot with the script as first input on the data |
||
36 | # on the second output |
||
37 | # the @file@ string in the gnuplot script is replaced by the data file. |
||
38 | |||
39 | script=$1; shift |
||
40 | data=`echo $1 | sed 's/\//\\\\\//g'`; shift |
||
41 | output=$1; shift |
||
42 | |||
43 | ( |
||
44 | echo "set terminal png" |
||
45 | echo "set output '$output'" |
||
46 | sed "s/@file@/$data/g" $script |
||
47 | ) > $script.plot |
||
48 | |||
49 | gnuplot $script.plot |
||
50 | rm $script.plot |
||
51 | </pre> |
||
52 | |||
53 | h2. Example |
||
54 | |||
55 | This operator is used plotting the results in the stream [[Data mining complexity]]. |