I want to detect the trend and seasonal changes of NDVI index. I want to know that for decomposing the time series, which one is the best, using the BFAST algorithm in R package or decomposing the time series in Matlab or excel or python?
There are all useful but i can recommend you to use either the time series in Matlab or excel.
There are a lot of packages about time series analysis in R. I suggest starting with the forecast package. The package's authors have published an excellent book about time series analysis that can be reached at https://otexts.com/fpp2/
I shared a sample code to decompose univariate time series data below:
#install.packages("forecast")
library(forecast)
data("gas")
# grafical approach
ggseasonplot(gas)
gglagplot(gas)
ggsubseriesplot(gas)
#first method to decompose time series
decomp1
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series