Pi-Change: Oil Price Application

The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

This vignette analyzes absolute daily changes in WTI spot oil prices from 2000 through 2009. The prior locations correspond to events used in the original application script: rising prices around 2003, Hurricane Katrina, Iran sanctions, and the 2008 global recession. This example illustrates how PI-Change can use geopolitical and economic event times as soft prior support for structural changes in the observed series.

library(PiChange)
library(ggplot2)

data("wti_oil")

event_dates <- as.Date(c("2003-01-01", "2005-08-26", "2006-12-23", "2008-07-01"))
penalty <- construct_penalty(
  time = wti_oil$date,
  centers = event_dates,
  width = 26 * 5,
  method = "mbic",
  family = "zag"
)
fit <- pi_change(wti_oil$abs_price_change, penalty, min_seg_len = 13 * 5)

change_indices <- changepoints(fit)
wti_oil[change_indices, c("date", "price", "abs_price_change")]
#>            date price abs_price_change
#> 1093 2004-05-17 41.53             0.11
#> 1954 2007-10-23 86.45             1.15
#> 2271 2009-01-27 41.67             4.83

plot(
  fit,
  ylab = "Absolute daily price change",
  title = "WTI oil price volatility with PI-Change estimates"
)

Here width = 26 * 5 represents 130 observation positions, approximately six months of trading days, while min_seg_len = 13 * 5 requires approximately one quarter of trading days per segment. These are application-specific choices, not universal defaults. The original dates, matched observation indices, and penalty values are retained in fit$penalty.

Reference

Jacobs, J. and Chen, S. (2026). Pi-Change: A Prior-Informed Multiple Change Point Detection Algorithm. https://doi.org/10.48550/arXiv.2605.01003.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.