Subscribe to DSC Newsletter

How to interpret a time series with the naked eye? What do you see in this graph?

In the example below, a few patterns are obvious, and don't require more than visual inspection for confirmation:

  • Periodicity: weekly
  • A spike at the begining, then after the launch, the traffic goes down, then significantly recovers a few months later (this is very typical)
  • The growth is super-linear (faster than linear) as the gaps between successive peaks are shrinking
  • We see more double peaks recently, and the weekly lows are getting higher
  • Since August, traffic outside US is growing faster than domestic traffic
If you can identify more patterns just with your naked eye, you are welcome to post what you see!

Views: 114

Replies to This Discussion

You can also see that there's a strong positive correlation between the two series
Yes, but the high for global traffic are much higher than for US traffic, like 1.5 times higher. Normal traffic for Global is 1.25 times higher than US.

Maybe spikes happen when a newsletter is sent, and the newsletter either reach more international people or international convert better.

By the way, where did you take your picture? Mine is a sunset on planet Mars.
It depends how you define high. There are some local int. high that are only slightly higher, and some local lows that are about 1.5 times higher as well. Vincent's last point explains this phenomenon, but positive correlation is another observation with naked eye independent of this.

Nice picture Sun, mine is actually a sunrise on planet Earth :-)
Traffice might not come down below 150 for short term. In long run it may cross 600 again and it may stay for some more time.


On Data Science Central

© 2020 is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service