features.load.RdExtract the features from the load profiles of each household. These features are prepared for identifying the prosumers and regular consumers.
features.load(loadsample, x.week, mor.be, aft.be)
| loadsample | A numeric matrix that presents the load profile of a household. It should be an 'm*48' matrix, in which 'm' demonstrates the number of days of the daily load profile that can be obtained. There need to be 48 observations of one single day load profile. |
|---|---|
| x.week | A vector of characters with 'm' elements that gives the day-of-the-week information of the 'm' days' load profile. Each element should be one from c('Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'). |
| mor.be | An integer presents the index of time stamp that PV generation has impact to the load profile. Taking the 48 half-hourly observations of both daily load and PV profiles as an example, if the PV generation has impact for the load profile at 06:00:00, then 'mor.be' should be chosen as 13 by counting the index number from 00:00:00. |
| aft.be | An integer presents the index of time stamp that PV generation usually vanished in a day. For example, if the PV generation vanished at 19:00:00, then 'aft.be' should be set as 39. |
output.features --- A numeric feature vector with 63 element extracted from the load profile of one household.
features.name --- A vector of characters that demonstrate the names of the features.
### Load the required packages. # library(matrixStats) # ### Import the load profiles. # data("loadsample") # ### Get the load profile matrix of the household X1. # loadsample.X1<-load.daily(loadsample$X1, load.date=loadsample$date, num.obs=48) # ### Compute the features. # features.X1<-features.load(loadsample.X1, unique(loadsample$week), mor.be=13, aft.be=39)