The new home of Azop's bitcoin stability and rainbow charts
Please clarify the meaning of “Seasonal”.
Seasonality refers to price movements that, on average, occur across an annual cycle. The overall impact is relatively low compared to the underlying trend and “random” component of the remainder.
I meant, how is it calculated? My guess is that you adjust the values of the seasonal chart such that it minimizes the (integral of the (difference between (seasonal chart + trend) and (actual price))); correct?
Also consider making another kind of seasonal chart based on the discrepancy from the long term linear (i.e. exponential on a log axis) regression line instead of from a moving average.
(By the way, I love your charts.)
It uses a technique called LOESS. You can give it a read here: https://stat.ethz.ch/R-manual/R-devel/library/stats/html/stl.html
Essentially, the trend is a fancy moving average and the seasonal pattern is typical movement for that day based on historic action. Both of these are calculated to minimise the remainder component.
On your other question, the value in brackets is the average daily growth rate of the trend line.
Your email address will not be published. Required fields are marked *
Save my name, email, and website in this browser for the next time I comment.