Search code examples
rfunctionstatisticsintercepttrend

R How to calculate the intercept of sen's slope function from trend package?

I want to estimate the precipitation trends (in percentage) at long-term. At the first time, I calculate the sen's slope from the "trend package" but it misses the intercept to apply this linear equation to find the predicted value of each year:

f(S) = QS + B (Q : sen's slope, S : parameter, B : intercept)

I use this example to explain my problematic:

# import data
require(trend)
data("maxau")
maxau

# Estimate Annual trend with sen's slope

s <- maxau[,"s"]
sens.slope(s) # equal to  -0.2876139 / year

Can you help me how to calculate the intercpt (B) of f(S) equation in order to calculate the predicted value of S parameter for each year ?

Thank you in advance for your help !

Solution

  • i had undestood that the R 'trend' package doesn't generate the intercept of Sen's slope

    you can try using the 'zyp' package which generates the intercept based on the median of the values formula.

    here is the documentation of the zyp package and this is an example with dataframe

    library(zyp)
df=data.frame(x=c(1,2,3,4,5),y=c(3,6,8,1,9))
zyp.sen(y~x,df)

    documentation: https://cran.r-project.org/web/packages/zyp/zyp.pdf