Problem with forecast se in the forecast package
On 28 December 2016 at 03:20, Adam Ginensky <adamno227 at gmail.com> wrote:
Hi, I don't know what's going on, but notice the 'bad' sd estimates are essentially double the 'correct' sd estimates. There is likely a connection.
Ouch! Got it. Thanks,
Ajay Shah ajayshah at mayin.org http://www.mayin.org/ajayshah http://ajayshahblog.blogspot.com [[alternative HTML version deleted]]