choice {kml} | R Documentation |
choice
lets the user choose the Clusterization
he wants
to export.
choice(x)
x |
[ClusterizLongData]: Contains all the clusterization found by KmL . |
choice
run in three steps:
Clusterization
Clusterization
Non applicable
choice
opens two graphic windows. In the left one, the Calinski
criteria of all Clusterization
found by kml
are represented.
They are filed according to the number of clusters, from the biggest to
the smallest. This windows let the user to select a
Clusterization
.
The selected Clusterization
is pointed with a dark spot.
In the right window, the Clusterization
selected by the user is represented.
When choice
is called, the Clusterization
having the larger Calinski criterion is selected.
It is possible to visualize the other Clusterization
by using the
arrows on the keyboard.
When the choice of a Clusterization
has been made and needs to be
exported, the use can go on to the next step by pressing "Return".
When a Clusterization
is being chosen, there are three possibilities :
Clusterization
on screen (but nothing else).
nomDeFichier
enables the user to export the Clusterization
.
Clusters are exported in the file
nomDeFichier-cluster.csv
. Criteria are exported in nomDeFichier-criteres.csv
. Distances and posterior probabilities are in nomDeFichier-distance.csv
(not implemented yet).
->
(like
->variablesNames
) enables the user to save the
Clusterization
. Clusters are stored in the variable
nomVariable_cluster.csv
as a data.frame
. Criteria are
saved in nomDeVariable_criters.csv
as lists. Distances and
posterior probabilities are saved nomDeVariable_distance.csv
(not implemented yet).
Finally, it is possible to export a graphic representation of the
Clusterization
. With the keyboard, the user can modify the aspect of the graph.
e
, r
and t
deal with individual trajectories.
d
, f
and g
deal with the average trajectories of clusters.
c
, v
and b
deal with subgroups.
h
, j
, k
, l
and o
deal with symbols on the graph.
More precisely:
When the final graph is choisen by pressing "Return". It can then be exported as usual with a right click on the figure.
Christophe Genolini
PSIGIAM: Paris Sud Innovation Group in Adolescent Mental Health
INSERM U669 / Maison de Solenn / Paris
Responsable : <genolini@u-paris10.fr>
Raphaël Ricaud
Laboratoire "Sport & Culture" / "Sports & Culture" Laboratory
University of Paris 10 / Nanterre
cld1 <- as.cld(gald()) kml(cld1,2:3,3) #choice(cld1)