Asymptotically (for high enough thresholds), the interarrival times (i.e. times between threshold crossings) follow a Mittag-Leffler distribution. The slider now chooses the maximum number of exceedances (the minimum height of the threshold). The estimates of the tail and (rescaled) scale parameters of the Mittag-Leffler distribution are plotted as the threshold varies from the minimum height to the 10th largest data point. To the left (high threshold), data are scarce and variance is high; to the right (low threshold), the asymptotics are off and bias is high. A region of 'stability' in the middle is choisen for a parameter estimate.
Tail parameter
to your choice of estimateFor instance, tail=0.85
and scale=3000
(days) seem to be a good fit
for the Bitcoin trade dataset.
This means that if the threshold is set to the k-th largest magnitude,
then the interarrival times are Mittag-Leffler distributed with parameters
tail = 0.85
and scale = 3000 / k^(1/tail)
.
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.