mvFUN: Running Mean-Variance Test.
In gmiskell/sensor_toolkit: Functions for Managing Sensor Networks

Description Usage Arguments Examples

This function gives the running MV test for an observation, relative to a proxy from some other observation/s.

1
2
3

mvFUN(x, obs, date, proxy, all.data = TRUE, intercept.theta = NA,
  slope.theta = NA, tau = NA, window.length = 72, truncate = FALSE,
  trun.direction = "over")

`obs`	The data under investigation.
`date`	A required column with the format 'YYYY-MM-DD HH:MM:SS'.
`proxy`	The comparison data.
`all.data`	This decides whether to use all provided data (all.data - 'TRUE'), or to use data from the latest day ('FALSE'), which makes the function possible to run as a schedule for real-time updating. Defaults to 'TRUE'.
`intercept.theta`	This sets the test threshold on whether to flag the data for the intercept. Defaults to 'NA' (no flags given).
`slope.theta`	This sets the test threshold on whether to flag the data for the slope. Defaults to 'NA' (no flags given).
`tau`	This sets the day threshold on whether to flag the data given consistent theta flags. Defaults to 'NA' (no flags given).
`window.length`	This defines the length of the sampled window, defined by row number. Default is 72 (three days of hourly data).
`truncate`	This is is subsets of the data should be derived, by using only values above or below the mean to derive the mean and variance.
`trun.direction`	This the side of the mean to take for the sample, if truncate is ‘TRUE'. Options are ’under' (below the sample mean) or 'over' (above the sample mean).