make.aggregation: Aggregate dataframe with a function.

In VicenteYago/sensorsUtils: SensorUtils

Description

Performs an aggregation within the interval defined by freq, by executing a provided function.

Usage

make.aggregation(df, freq, fun)

Arguments

df	a dataframe, the first column must be the time index.
freq	frequency of the observations in the dataframe, same as ts object.
fun	a function like mean(), max(), etc.

Details

For security reasons the dataframe must have a number of observations multiple of freq, also all the time index must be complete i.e without silents losts, see fortify_df for achieve this.

Its desirable that the time interval present in the input dataframe include complete periods of time. For example if the time index (first column), takes the range ['2019/01/01 00:00:00', ..., '2019/01/03 20:00:00'] will be more convenient to modify it to ['2019/01/01 00:00:00', ..., '2019/01/03 23:00:00'] for obtain the last day completely, that effect can be acomplish with the function fortify_df, using the parameter fInterval.

Author(s)

JV Yago

Examples

# Mean of 3 day data by days
a = as.POSIXct("2019/01/01", tz = "UTC")
b = as.POSIXct("2019/01/03 23:00:00", tz = "UTC")
df <- data.frame(dates = seq.POSIXt(a,b,"hour"), A = runif(24*3))
make.aggregation(df, freq = 24, mean)

# The same with multicolumn datadrame
a = as.POSIXct("2019/01/01", tz = "UTC")
b = as.POSIXct("2019/01/03 23:00:00", tz = "UTC")
df<-data.frame(dates = seq.POSIXt(a,b,"hour"),
              A = runif(24*3),
              B = rep(1:3,each=24))
make.aggregation(df, freq = 24, mean)

# The previus example but using a 12 hour mean
a = as.POSIXct("2019/01/01", tz = "UTC")
b = as.POSIXct("2019/01/03 23:00:00", tz = "UTC")
df<-data.frame(dates = seq.POSIXt(a,b,"hour"),
              A = runif(24*3),
              B = rep(1:3,each=24))
make.aggregation(df, freq = 12, mean)

VicenteYago/sensorsUtils documentation built on Jan. 29, 2022, 12:44 a.m.