TraMineR (version 1.1)

seqefsub: Searching for frequent subsequences

Description

Return a list of frequent subsequences satisfying a minimum support. Several time constraints can be set to restrict the search to specific time periods or subsequences durations.

Usage

seqefsub(seq, strsubseq = NULL, minSupport = NULL, pMinSupport = NULL, 
  constraint = seqeconstraint(), maxK = -1)

Arguments

seq

A list of event sequences

strsubseq

Can be used to look for specific subsequences. See details.

minSupport

The minimum support (in number of sequence)

pMinSupport

The minimum support (in percentage, will be rounded)

constraint

Time constraint object, use seqeconstraint

maxK

The maximum number of event in a subsequence

Value

• A subseqelist object which contain at least the following objects:
• subseq
  A list of subsequence (event sequence object) as a seqelist.
• data
  A data.frame with subsequence specific information such as the support
• constraint
  The constraint object used to compute the subsequence.
• count
  The number of sequences.
• subseqelist can be selected using the [] operator (see example). There are print and plot methods for subsequelist

Details

The support is counted per sequence and not per occurence. The support can be set through pMinSupport as a percentage (between 0 and 1 and it will be rounded), or throught minSupport as number of sequence. It is possible to specify time constraints using constraint argument (see seqeconstraint function). The strsubseq parameter can be used to loof for specific (user specified) subsequences. The format is the same as the one used to display subsequences (see str.seqelist). Each group of event should be enclosed in brace () and separeted with comma ,. A - can be used to specify a time gap. For instance "(FullTime)-(PartTime, Children)" will look for the subsequence "FullTime" event followed by the events "PartTime" and "Children" appearing at the same time.

See Also

See plot.subseqelist to plot the result. See seqecreate for creating event sequences. See seqeapplysub to count the number of occurence of frequent subsequences in each sequence. See is.seqelist about seqelist.

Examples

data(actcal.tse)
actcal.seqe <- seqecreate(actcal.tse)

##Searching for frequent subsequences, that is, appearing at least 20 times
fsubseq <- seqefsub(actcal.seqe, minSupport=20)
##The same using a percentage
fsubseq <- seqefsub(actcal.seqe, pMinSupport=0.01)
##Getting a string representation of subsequences
##Ten first subsequences
fsubseq[1:10]

##Using time constraints
##Looking for subsequence starting in summer (between june and september)
fsubseq <- seqefsub(actcal.seqe, minSupport=10, 
  constraint=seqeconstraint(ageMin=6, ageMax=9))
fsubseq[1:10]

##Looking for subsequence contained in summer (between june and september)
fsubseq <- seqefsub(actcal.seqe, minSupport=10, 
  constraint=seqeconstraint(ageMin=6, ageMax=9, ageMaxEnd=9))
fsubseq[1:10]

##Looking for subsequence enclosed in a 6 month period
## and with a maximum gap of 2 month
fsubseq <- seqefsub(actcal.seqe, minSupport=10, 
  constraint=seqeconstraint(maxGap=2, windowSize=6))
fsubseq[1:10]