library(zeitgebr)
library(ggetho)
# We make toy data
data(dams_sample)
dt <- copy(dams_sample)
dt[, moving:= activity>0]
#1min, native
per_dt <- periodogram(moving, dt, FUN = chi_sq_periodogram)
per_dt <- find_peaks(per_dt)
out <- ggperio(per_dt, aes(y=power, colour=period_group, peak=peak))
out + geom_line() + facet_wrap(~id)
# downsampling to 10min. works well
per_dt <- periodogram(moving, dt, FUN = chi_sq_periodogram, resample_rate =1/mins(10))
per_dt <- find_peaks(per_dt)
out <- ggperio(per_dt, aes(y=power, colour=period_group, peak=peak))
out + geom_line() + facet_wrap(~id)
# downsampling to 1 read per 10s (6x). artefacts in periodogram that were great before!
per_dt <- periodogram(moving, dt, FUN = chi_sq_periodogram, resample_rate =1/10)
per_dt <- find_peaks(per_dt)
out <- ggperio(per_dt, aes(y=power, colour=period_group, peak=peak))
out + geom_line() + facet_wrap(~id)
