Wednesday, March 7, 2012

Forecasting Flowers

Last March I had a post that I titled "Modelling Flowers on the Wall". It was to do with the Great Victoria Flower Count that the people of this town engage early in each Spring in the hope of annoying friends and family living in less temperate parts of our continent.

In that post I developed an extremely sophisticated forecasting model that can be used to predict the flower count each year. The 2012 count is now on, so before the final numbers are in I'd better give you my forecast!

Here it is: 1,573,026,000 blooming flowers.

Oh, I almost forgot - the forecast standard error is 2,451,832,000. I think I'll be O.K.!

(You'll find the data and EViews workfile I used here.)

Postscript (9 March, 2012): The official tally was announced yesterday - 2,043,380,013.



© 2012, David E. Giles

3 comments:

  1. Prof.Giles this is also off topic but I wanted to know if i have an MA(2) process what should the invertibility frontier be 2? Please Help , a caribbean student.

    ReplyDelete
  2. See http://web.uvic.ca/~dgiles/blog/MA2.pdf

    ReplyDelete
  3. Thanks so much!

    ReplyDelete