anomaly (version 4.0.2)

capa.uv: Detection of univariate anomalous segments and points using CAPA.

Description

A technique for detecting anomalous segments and points in univariate time series data based on CAPA (Collective And Point Anomalies) by Fisch et al. (2018). CAPA assumes that the data has a certain mean and variance for most time points and detects segments in which the mean and/or variance deviates from the typical mean and variance as collective anomalies. It also detects point outliers and returns a measure of strength for the changes in mean and variance. If the number of anomalous windows scales linearly with the number of data points, CAPA scales linearly with the number of data points. At worst, if there are no anomalies at all and max_seg_len is unspecified, the computational cost of CAPA scales quadratically with the number of data points.

Usage

capa.uv(
  x,
  beta = NULL,
  beta_tilde = NULL,
  type = "meanvar",
  min_seg_len = 10,
  max_seg_len = Inf,
  transform = robustscale
)

Arguments

x

A numeric vector containing the data which is to be inspected. The time series data classes ts, xts, and zoo are also supported.

beta

A numeric vector of length 1 or max_seg_len - min_seg_len + 1 indicating the penalty for adding additional collective anomalies of all possible lengths. If an argument of length 1 is provided the same penalty is used for all collective anomalies irrespective of their length. The default value is 4log(n), where n denotes the number of observations.

beta_tilde

A numeric constant indicating the penalty for adding an additional point anomaly. It defaults to 3log(n), where n denotes the number of observations.

type

A string indicating which type of deviations from the baseline are considered. Can be "meanvar" for collective anomalies characterised by joint changes in mean and variance (the default), "mean" for collective anomalies characterised by changes in mean only, or "robustmean" for collective anomalies characterised by changes in mean only which can be polluted by outliers.

min_seg_len

An integer indicating the minimum length of epidemic changes. It must be at least 2 and defaults to 10.

max_seg_len

An integer indicating the maximum length of epidemic changes. It must be at least the min_seg_len and defaults to Inf.

transform

A function used to transform the data prior to analysis by capa.uv. This can, for example, be used to compensate for the effects of autocorrelation in the data. Importantly, the untransformed data remains available for post processing results obtained using capa.uv. The package includes several methods that are commonly used for the transform, (see robustscale and ac_corrected), but a user defined function can be specified. The default values is transform=robust_scale.

Value

An instance of an S4 class of type capa.uv.class.

References

2018arXiv180601947Fanomaly

Examples

# NOT RUN {
library(anomaly)
data(machinetemp)
attach(machinetemp)
res<-capa.uv(temperature,type="mean")
canoms<-collective_anomalies(res)
dim(canoms)[1] # over fitted due to autocorrelation
psi<-0.98 # computed using covRob
inflated_penalty<-3*(1+psi)/(1-psi)*log(length(temperature))
res<-capa.uv(temperature,type="mean",beta=inflated_penalty,
             beta_tilde=inflated_penalty)
summary(res)
plot(res)

library(anomaly)
data(Lightcurves)
### Plot the data for Kepler 10965588: No transit apparent
plot(Lightcurves$Kepler10965588$Day,Lightcurves$Kepler10965588$Brightness,xlab = "Day",pch=".")
### Examine a period of 62.9 days for Kepler 10965588
binned_data = period_average(Lightcurves$Kepler10965588,62.9)
inferred_anomalies = capa.uv(binned_data)
plot(inferred_anomalies)

# }