oil_price_unemployment.Rmd

Structural Break for Oil Price Analysis
=======================================================

```{r, echo=FALSE}
install.packages("changepoint")
require(changepoint)

setwd("/Users/Yue/Documents/meetup_doc/TorontoMachineLearningBookClub/data")

# import data
data_oil = read.csv("data_oil.csv")
data_oil = data_oil[-nrow(data_oil),]

# convert to time series data frame
value = as.numeric(as.character(data_oil$VALUE))

oil_ts = ts(value, frequency=12, start=c(2000,1))

plot(oil_ts)

meanvalue.PELT = cpt.mean(oil_ts, method="PELT") # mean chanepoints using PELT

cpts(meanvalue.PELT) # list the changepoints. All observatioins are changes

plot(meanvalue.PELT)

meanvalue.BinSeg = cpt.mean(oil_ts, method="BinSeg") # need to increase Q to increase changepoints
cpts(meanvalue.BinSeg)

plot(meanvalue.BinSeg)

# detect change in variance

variancevalue =cpt.var(diff(oil_ts), method="PELT")

cpts(variancevalue)

plot(variancevalue)

par(mfrow=c(2,1))

plot(oil_ts)
plot(variancevalue)

variance_n_value = cpt.var(diff(oil_ts), method="SegNeigh", Q=6) # maxiumu 6 number of changepoints

cpts(variance_n_value)

# decompose time series into trend, seasonality and random component

d = decompose(oil_ts)
plot(d)
```