0
$\begingroup$

Out of my curiosity, when trying to learn forecasting from fpp3, I found that I can't extract seasonal info from global_economy but I can't explain why.

My assumption is that the dataset doesn't have seasonal info to extract. I briefly see that there not much seasonality in the dataset using autoplot.

But I could be wrong. Any suggestions?

# A dable:           15,150 x 7 [1Y]
# Key:               Country, .model [263]
# STL Decomposition: GDP = trend + remainder
   Country     .model    Year         GDP       trend   remainder season_adjust
   <fct>       <chr>    <dbl>       <dbl>       <dbl>       <dbl>         <dbl>
 1 Afghanistan STL(GDP)  1960  537777811.  467453507.   70324304.    537777811.
 2 Afghanistan STL(GDP)  1961  548888896.  571296978.  -22408082.    548888896.
 3 Afghanistan STL(GDP)  1962  546666678.  675140449. -128473771.    546666678.
 4 Afghanistan STL(GDP)  1963  751111191.  800103336.  -48992145.    751111191.
 5 Afghanistan STL(GDP)  1964  800000044.  932454256. -132454212.    800000044.
 6 Afghanistan STL(GDP)  1965 1006666638. 1070667576.  -64000938.   1006666638.
 7 Afghanistan STL(GDP)  1966 1399999967. 1217560744.  182439222.   1399999967.
 8 Afghanistan STL(GDP)  1967 1673333418. 1353217120.  320116298.   1673333418.
 9 Afghanistan STL(GDP)  1968 1373333367. 1453277699.  -79944332.   1373333367.
10 Afghanistan STL(GDP)  1969 1408888922. 1544685304. -135796382.   1408888922.
$\endgroup$
2
  • 2
    $\begingroup$ What you show us is annual data, so I think you've answered your own question, except that "not much seasonality" is not good wording for "zero information on seasonality". $\endgroup$
    – Nick Cox
    Commented Apr 25, 2020 at 10:12
  • $\begingroup$ @NickCox silly me! it should not have any seasonality at all because it is an annual data. Thanks! $\endgroup$ Commented Apr 25, 2020 at 12:49

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.