removeVisits: Remove clinical visits with no history of length 'maxgap'.

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

Description

Remove clinical visits with no history of length maxgap.

Usage

removeVisits(data, maxgap, maxlength, tType, save = TRUE, outDir)

Arguments

data	dataframe, see Details
tType	char, tumor volume type nickname, see getVolTypeName
save	bool, save removed data?
outDir	char, full path to output directory

Details

clinical visits that do not have a history of maxgap time cannot be evaluated (such vists would have indeterminate values for temporal patterns item1 –> item2 when the maxgap exisits between these items) e.g., at the first observed visit for a patient, I search for item2's existence, and then look backwards in time to see if item1 happend within the maxgap. I wouldn't be able to determien this pattern's value because no visits are before the first visit. Therefore, such clinical visits should be removed.

data dataframe, rows are clinical visits to be classified, columns are features including clinical visits ids and at least the following columns:

• id char, clinical visit id in the format "patientID.eventID"

• iois char, patient id

• eventID char, event id

Examples

data("features_ratechange_sup0.4g60l2z2") # features and labels for each clinical visit
maxgap <- 60 # constants used to create features
maxlen <- 2
t <- 'rate'

# format
names <- colnames(feats)
feats <- data.frame(id=row.names(feats),feats)
colnames(feats) <- c('id',names)
cat('...overall samples: ', nrow(feats), '\n')
feats <- prepLaterality(feats) 
feats <- prepLocation(feats)

cat('...removing clinical visits with no history of length maxgap:', maxgap, '\n')
feats <- removeVisits(feats, 
                     maxgap=maxgap, 
                     maxlength=maxlen, 
                     tType=t, 
                     save=F, 
                     outDir=NA)

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