This package performs time series analysis on a set of data that may be highly correlated. The estimator performs a “Blocking Method” to output the true mean and standard error (variance) for a given column of data. For more information on the method refernce:
@article{article,
author = {Flyvbjerg, Henrik and Petersen, H.G.},
year = {1989},
month = {07},
pages = {},
title = {Error Estimates on Averages of Correlated Data},
volume = {91},
journal = {The Journal of Chemical Physics},
doi = {10.1063/1.457480}
}The function estimate returns two values: the mean and variance. Insert a
Vector of Float64 to recieve the statistics of that column.
# using the package
using BlockingMethod
# set of data
x = rand(1000)
mn, ste = estimate(x)