MENU
R'MES general view rmes.gaussien rmes.poisson.composee rmes.poisson
rmes.gfam rmes.format rmes.histo rmes.compar
rmes.pyramide


Splus Function: rmes.pyramide
Graphical Display of Results from R'MES Programs.



Description Usage Required Arguments Optional Arguments Note Details Examples

Description

The Splus function rmes.pyramide displays simultaneously the statistics of words of a given length and their sub-words in a pyramid of gray-level. The graphical display is a PostScript file.

NOTE: Pyramids of word families cannot be done.

Usage


rmes.pyramide(prefix, dessin, phase=0, lg.mot, nb.mot,
              neg=F, debut=1, rang=F, nb.col, lang="fr")

Required Arguments
prefix
Character string equal to the pathname of the file containing the information to be displayed. This file should have been created by one of the following commands: rmes.gaussien, rmes.poisson, rmes.poisson.composee.
BE CAREFUL: the suffix (.0, .1, .2, .3 or .4) must not be mentionned.
dessin
Character string equal to the pathname of the PostScript file where the graphics will be stored.
BE CAREFUL: if this file exists, it will be replaced.

Sommaire

Optional Arguments

phase
Integer between 0 and 4 equal to the suffix that must be added to prefix according to the phase of the words to be considered.
When the words are studied on a particular phase (phase=1, 2 or 3), all the files prefix.i for i=1 to 3 must be available. By default, 0.
lg.mot
Length of words that will be represented. It must be less or equal 8. By default, the maximal length of words (lg.max) analyzed in the file prefix.phase.
nb.mot
Number of words (pyramids) that will be represented. By default, pyramids of all the words of length lg.mot.
neg
Logical value. When TRUE, words with a negative statistic are first considered (their statistics will increase). By default, the order of the words corresponds to the decreasing order of their statistics.
debut
Rank of the first word to be represented according to the order defined by the argument neg. By default, the first pyramid corresponds to the most exceptional word.
rang
Logical value. When TRUE, the gray-level depends on the rank of the statistics. By default, it depends on the statistic values.
nb.col
Integer equal to the number of pyramids on each line of the display.
By default, it is calculated from the pyramid size.
lang
When equal to "fr", labels and messages are in french. When equal to "en", labels and messages are in english. By default, they are in french.

Sommaire

Note

In this version, the environment variable $RMES_DIR must be set to the pathname of the installation directory of the system.

Details

 A word of length lg.mot is represented by a pyramid of lg.mot-lg.min+1 stages, where lg.min is the minimal word length analyzed in the file prefix.phase. Let us suppose lg.min=3. The last stage is composed of a unique square colored according to the value of the statistic (or to the value of its rank when the argument rang is TRUE). The stage beneath is made of two squares corresponding to the 2 sub-words of length lg.mot-1 of the word. They are colored according to the value of their statistics (or to their ranks when the argument rang is TRUE). And so on, to the first stage made of lg.mot-2 squares corresponding to the lg.mot-2 sub-words of length 3 of the word.

For a given length, the words of null statistics are colored in white. The words with maximal statistics (positive and negative) are colored in black. A uniform gray-level is used between these extrema. Each stage has its own gray-level.

To distinguish the negative statistics from the positive ones, the sign of the statistic of the upper word is put above each pyramid and the symbol O is added in the middle of the squares that correspond to sub-words with statistic of opposite sign.
The statistic of the sub-word that could not be calculated is symbolized by "x" in the corresponding square.

Sommaire

Examples

1- Pyramids of 6-words (starting by the most under-represented one) in the ecomori sequence under maximal models. The plot is stored in the PostScript file ecomori-pyr.ps and viewed by "ghostview" (205 pages). The file corresponding to the input prefix ecomori.gaussien.mart.3-6_max have been previously obtained by rmes.gaussien (example 4 in rmes.gaussien). 

> rmes.pyramide("ecomori.gaussien.mart.3-6_max", 
                 dessin="ecomori-pyr.ps",
                 lg.mot=6, neg=T, nb.col=4)
2- Pyramids of the 20 6-words with smallest statistics (negative) from the 10th smallest value. The previous file ecomori-pyr.ps will be replaced. 

> rmes.pyramide("ecomori.gaussien.mart.3-6_max", dessin="ecomori-pyr.ps",
	         lg.mot=6, neg=T, nb.mot=20, debut=10, lang="en")
3- Pyramids of the most over-represented 6-words in phase 3 of a coding DNA sequence from E. coli (coli-codant) under M2_3. The plot is stored in the PostScript file coli.6.pyr3.ps. The four files corresponding to the input prefix coli-codant.gaussien.4-6_2 have been previously obtained by rmes.gaussien (see example 5 in rmes.gaussien). 

> rmes.pyramide("coli-codant.gaussien.4-6_2", phase=3, lg.mot=6,
                 dessin="coli.6.pyr3.ps", nb.col=3, nb.mot=12)

Sommaire