seqdiff {TraMineR} | R Documentation |
Decompose the difference between groups of sequences
seqdiff(seqdata, group, cmprange = c(0, 1), seqdist_arg=list(method="LCS",norm=TRUE))
seqdata |
The sequence to analyse |
group |
The group variable |
cmprange |
The range used to compare subsequences |
seqdist_arg |
argument passed directly to seqdist as a list |
Analyses at each timestamp the sequence discrepancy within a sliding time window (of range defined by cmprange
)
that is explained by the group
variable.
The method computes a distance matrix, using seqdist
at each timestamp and then derives the explained
discrepancy with dissassoc
.
There are print and plot methods for the result returned.
A seqdiff
object, with the following items:
stat |
A data.frame with three statistics (PseudoF, PseudoR2 and PseudoT) for each timestamp of the sequence,
see dissassoc |
variance |
A data.frame with, at each time stamp, the discrepancy within each group defined by the
group variable and for the whole population. |
Studer, M., G. Ritschard, A. Gabadinho, and N. S. Müller (2009) Discrepancy analysis of complex objects using dissimilarities. In H. Briand, F. Guillet, G. Ritschard, and D. A. Zighed (Eds.), Advances in Knowledge Discovery and Management, Studies in Computational Intelligence. Berlin: Springer.
Studer, M., G. Ritschard, A. Gabadinho and N. S. Müller (2009) Analyse de dissimilarités par arbre d'induction. In EGC 2009, Revue des Nouvelles Technologies de l'Information, Vol. E-15, pp. 7-18.
dissassoc
to analyse the association with the whole sequence
## Defining a state sequence object data(mvad) mvad.seq <- seqdef(mvad[, 17:86]) ## Building dissimilarities mvad.diff <- seqdiff(mvad.seq, group=mvad$gcse5eq) print(mvad.diff) plot(mvad.diff) plot(mvad.diff, stat="Variance")