generate_features_dataset: Generate time series used in "Feature Extraction"-chapter of...

Description Usage Arguments Value Examples

Description

Generate time series used in "Feature Extraction"-chapter of the manual.

Usage

1
2
3
4
generate_features_dataset(seed = NULL, n_measurements_per_day = 5,
  n_days = 21, m = c(3, 6), sdev = c(0.5, 0.5), ampl = c(1, 4),
  range = c(0, 10), missingness_prob = c(0.1, 0.5),
  autocorrelation = 0.6, plot = FALSE)

Arguments

seed

random seed.

n_measurements_per_day

number of data points per day.

n_days

assessment period (in days).

m

mean at start (first element) and end (second element) of s.

sdev

sd at start (first element) and end (second element) of s.

ampl

amplitude of circadian rhythms (1-day, 7-day period) in s.

range

min, max range of s.

missingness_prob

minimal and maximal missingness probability of s.

autocorrelation

strength of autocorrelation component in s.

plot

if TRUE, the series is plotted (s ~ t). Default is FALSE.

Value

a data.frame, containing the generated timeseries, with time in column t and the signal in column s.

Examples

1
2
3
4
5
d <- generate_features_dataset(seed = 123, plot = TRUE)
head(d)

fm <- lm(s ~ t, d)
summary(fm)

jruwaard/emaph documentation built on May 3, 2019, 8:04 p.m.