Smisc (version 0.3.9.1)

timeIntegration: Approximate the integral of a vector of data over time

Description

Integrate a series over time by calculating the area under the "curve" of the linear interpolation of the series (akin to the Trapezoid rule). This is especially useful in calculating energy usage: kilowatt-hours, watt-seconds, etc.

Usage

timeIntegration(data, time = names(data), lower = time[1],
  upper = time[length(time)], check.plot = FALSE, units = c("hours",
  "minutes", "seconds"))

Arguments

data

Vector of numerical data

time

Vector of timestamps which correspond to data. These can either character or POSIXct.

lower

The time (character or POSIXct) of the lower bound of the integration

upper

The time (character or POSIXct) of the upper bound of the integration

check.plot

=TRUE makes a plot which illustrates the integration.

units

The units of integration, defaults to hours. It is only required to supply enough characters to uniquely complete the name.

Value

The approximation of the integral by joining the points in the series in a linear fashion and calculating the area under this "curve".

Details

If upper or lower does not correspond to a data point, a linear interpolation is made between the two neighboring time points to predict the resulting data value.

Examples

# NOT RUN {
# Some example power data
data(PowerData)

par(mfrow = c(2, 1))

# Calculate the kilowatt-minutes, display graph which shows how the
# integration is done.  This example calculates the integral using
# a contiguous subset of the data
int1 <- timeIntegration(PowerData,
                       # Convert to POSIXct in order to subtract time
                       lower = "5/6/2008 17:00:09",
                       upper = "5/6/2008 17:01:36",
                       check.plot = TRUE, units = "m")

# This example calculates the integral for all the data in 'powerData'
int2 <- timeIntegration(PowerData, check.plot = TRUE, units = "m")

# Print the outcome
pvar(int1, int2)
# }