Observe how, thanks to the ordering, most “hot” (red) values representing high
similarities are located close to the diagonal. The diagonal itself is trivial.
Finally, it may be useful to explore the clusters’ species contents directly, i.e. to
reorder the original data table according to the group memberships. In order to avoid
large values to clog the picture, Jari Oksanen proposes vegemite(), a function of
vegan that can use external information to reorder and display a site-by-species
data table where abundance values can be recoded in different ways. If no specific
order is provided for the species, these are ordered by their weighted averages on the
site scores. Note that the abundance scale accepted by vegemite() must be made
of one-digit counts only (e.g. from 0 to 9). Any two- or more digit value will cause
vegemite() to stop. Our abundance data, expressed on a 0–5 scale, cause no
problem. Other data may have to be recoded. vegemite() itself makes an internal
use of another function, coverscale(), devoted to the rescaling of
vegetation data.
# Ordered community table
# Species are ordered by their weighted averages on site scores.
# Dots represent absences.
or <- vegemite(spe, spe.chwo)
32222222222 111111
1111
09876210543959876506473221341
Icme 5432121......................
Abbr 54332431.....1...............
Blbj 54542432.1...1...............
Anan 54432222.....111.............
Gyce 5555443212...11..............
Scer 522112221...21...............
Cyca 53421321.....1111............
Rham 55432333.....221.............
Legi 35432322.1...1111............
Alal 55555555352..322.............
Chna 12111322.1...211.............
Titi 53453444...1321111.21........
Ruru 55554555121455221..1.........
Albi 53111123.....2341............
Baba 35342544.....23322.........1.
Eslu 453423321...41111..12.1....1.
Gogo 5544355421..242122111......1.
Pefl 54211432....41321..12........
Pato 2211.222.....3344............
Sqce 3443242312152132232211..11.1.
Lele 332213221...52235321.1.......
Babl .1111112...32534554555534124.
Teso .1...........11254........23.
Phph .1....11...13334344454544455.
Cogo ..............1123......2123.
Satr .1..........2.123413455553553
Thth .1............11.2......2134.
sites species
29
27
94
4 Cluster Analysis
similarities are located close to the diagonal. The diagonal itself is trivial.
Finally, it may be useful to explore the clusters’ species contents directly, i.e. to
reorder the original data table according to the group memberships. In order to avoid
large values to clog the picture, Jari Oksanen proposes vegemite(), a function of
vegan that can use external information to reorder and display a site-by-species
data table where abundance values can be recoded in different ways. If no specific
order is provided for the species, these are ordered by their weighted averages on the
site scores. Note that the abundance scale accepted by vegemite() must be made
of one-digit counts only (e.g. from 0 to 9). Any two- or more digit value will cause
vegemite() to stop. Our abundance data, expressed on a 0–5 scale, cause no
problem. Other data may have to be recoded. vegemite() itself makes an internal
use of another function, coverscale(), devoted to the rescaling of
vegetation data.
# Ordered community table
# Species are ordered by their weighted averages on site scores.
# Dots represent absences.
or <- vegemite(spe, spe.chwo)
32222222222 111111
1111
09876210543959876506473221341
Icme 5432121......................
Abbr 54332431.....1...............
Blbj 54542432.1...1...............
Anan 54432222.....111.............
Gyce 5555443212...11..............
Scer 522112221...21...............
Cyca 53421321.....1111............
Rham 55432333.....221.............
Legi 35432322.1...1111............
Alal 55555555352..322.............
Chna 12111322.1...211.............
Titi 53453444...1321111.21........
Ruru 55554555121455221..1.........
Albi 53111123.....2341............
Baba 35342544.....23322.........1.
Eslu 453423321...41111..12.1....1.
Gogo 5544355421..242122111......1.
Pefl 54211432....41321..12........
Pato 2211.222.....3344............
Sqce 3443242312152132232211..11.1.
Lele 332213221...52235321.1.......
Babl .1111112...32534554555534124.
Teso .1...........11254........23.
Phph .1....11...13334344454544455.
Cogo ..............1123......2123.
Satr .1..........2.123413455553553
Thth .1............11.2......2134.
sites species
29
27
94
4 Cluster Analysis