runSPM: Run sequential pattern mining

In novasmedley/gbmSpm: Find predictive sequential patterns for clinical glioblastoma patients

Description

Run sequential pattern mining

Usage

runSPM(event, suppList, maxgap, maxlen, maxsize, tType, outputDir, createT)

Arguments

event	dataframe of events, where rows are single events, and least the following columns: iois patient id eventID event id, see resetEventIDsToDays eventName unique event name, see createEventNames
suppList	one or more supports to be used as a parameter, see slot in codeSPparameter-class
maxgap	see slot in SPparameter-class
maxlen	see slot in SPparameter-class
maxsize	see slot in SPparameter-class
tType	char, shortcut for type of tumor volume information
outputDir	full path to output folder, e.g., where to save spm results
createT	logical, create or load previously created transactions

Details

This will automatically create three types of files:

• transactions - returned from createTransactions, saved in .txt

• patterns - returned from findPatternDays, saved in .rds

• features - returned from getFeatureVectors, saved in .rds

See Also

getSeqPatterns, getVolTypeName

Examples

data("fake_data")
outputDir <- '~/test'

# save tumor location and laterility strings before event cleaning
fake_tumorInfo <- fake_data$events 
fake_demo <- fake_data$demo

# clean data
fake_data$events <- cleanData(fake_data$events, tType = 'rate')
cat('...',nrow(fake_data$events), " events left for SPM after cleaning", '\n')

# collect patient info for each event
fake_data <- merge(fake_data$events, fake_data$person, by='iois', all.x=T) 

# prep for each event, since age does change
# get survival labels, these also change
fake_data <- prepDemographics(fake_data, fake_demo) 
fake_data <- prepSurvivalLabels(fake_data) 

# get first tumor location
fake_data <- getTumorLocation(fake_data, fake_tumorInfo) 

runSPM(event=fake_data,
       suppList=suppList, 
       maxgap=60, 
       maxlen=2, 
       maxsize=2, 
       tType='rate', 
       outputDir='~/test')

novasmedley/gbmSpm documentation built on May 17, 2019, 10:39 a.m.