windowFeatures: Time Window Aggregation
In LoneWolf6/UJ-Analysis: Creation and Analysis of User Journey Data

Description Usage Arguments Value See Also Examples

This function creates time-based minimum, maximum, mean, sum, and standard deviation of numeric columns of a data frame. These values are calculated based on specific time steps provided by the user.

1	TimeWindow(input, parallel, cores, target, no.use, steps)

`input`	an object of type data.frame. Preferably the output of reshapeData().
`parallel`	an optional logical value True or False. Parameter specifies if parallel computation is enabled according to the foreach package. Can result in tremendous improvement of computation time. Default is False.
`cores`	an optional non-negative integer. The desired number of cores utilized if parallel = T. Default is the number of cores available -1.
`target`	a mandatory character string defining the dependent variable which is therefore not included in the computation.
`no.use`	an optional character string that defines other variables that should not be used.
`steps`	an optional non-negative numeric value defining the time window. Steps defines how many previous points in time are considered (including current point in time). Default is 2.

An object of type data.frame in the same form as data including additional features.

foreach, registerDoSNOW, makeCluster, detectCores

# create data frame with mandatory columns
data = data.frame('id'=rep(c(1:5), each=8),
      'type'=rep(c('Var1', 'Var2'), times=20),
      'value'=rep(c(1:5), times=8),
      'date'=rep(seq(as.Date("2000/1/1"), by = "day", length.out=20), each=2))

# create rectangle version of user journey
dat = reshapeData(data)

# use function to create new features based on time windows
dat_feat = TimeWindow(dat, parallel=T, cores=2, target='Var1', no.use=F, steps=2)