CODEWITHSUNDEEP
r
cgrhodes
cgrhodes

Reputation: 11

De-trending a dataset

I am trying to complete an autoregression on my dataset, examining the relationship between winter temperatures and deer populations. I ran an Augmented Dickey Fuller Test and realized that I need to de-trend my data before completing an autoregression.

Does anyone know how to de-trend a time series without losing the years?

Here is a subset of the data, if that helps.

I've tried to de-trended my deer population data set before, but was unable to preserve year. 

print(deer_pop)

   year population
1  1976     184729
2  1977     181017
3  1978     163250
4  1979     160169
5  1980     214924
6  1981     198624
7  1982     166286
8  1983     169222
9  1984     175300
10 1985     204395
11 1986     206772
12 1987     198760
13 1988     229226
14 1989     226091
15 1990     198285
16 1991     220106
17 1992     215492
18 1993     216814
19 1994     207537
20 1995     233524
21 1996     255604
22 1997     254299
23 1998     292072
24 1999     331435
25 2000     291474

Upvotes: 1

Views: 46

Answers (1)

eastclintw00d
eastclintw00d

Reputation: 2364

This is a question that you should rather ask at Cross Validated. A simple method for detrending data is to take the first difference, i.e., to substract the population at time t - 1 from the population at time t. Use:

diff(deer_pop$population)

You will loose one observation by doing so. You may need to take higher order differences, in order to get rid of the trend. This means you will loose more observations.

Upvotes: 1

Related Questions