time series introduction

[josef...@gmail.com]

I haven't been keeping up with many things,

Checking some twitter guys, I found a link to this
http://www.blackarbs.com/blog/time-series-analysis-in-python-linear-models-to-garch/11/1/2016

and I found this similar page from a 2014 course
http://iacs-courses.seas.harvard.edu/courses/am207/blog/lecture-17.html

and a potentially useful dataset is
http://archive.ics.uci.edu/ml/datasets/Individual+household+electric+power+consumption

very messy minute data with missing values, outliers or outlier episodes (vacation?)

and 3 seasons, daily, weekly and annual.

Because it's individual household data, it's very noisy. Aggregate data is much nicer because the law of large numbers and central limit theorems kick in and smooth the individual noise.

(And it's a good dataset to struggle with pandas.)

Josef