Implementation of algorithm from Robust peak detection algorithm in Python.
This algorithm has been modified in order to work with real-time streams. It doesn't recalculate all data points on arrival of a new one.
I used this algorithm with a watch ticks dataset, taken from a phone microphone.
For semplicity, lag is the same size of the first dataset quartz1
, that is used to calculate the mean, the variance and the standard deviation of the first 1280 rows.
After the initialization, the stream has been simulated with the quartz2
dataset; for every element, the mean, the variance and the standard deviation is recalculate.