Skip to content

First functional version of Chhabra-Jensen method to calculate the Multifractal spectrum of a time series

Latest
Compare
Choose a tag to compare
@cndesantana cndesantana released this 30 Dec 21:21
· 32 commits to master since this release

This approach, presented by Chhabra & Jensen (1989) [1], determines the multifractal spectra directly from the signal without using a Legendre transform.

To run an example of this method, you can do the following:

    using Multifractal

    function main(inputfile::ASCIIString, extensionDq::ASCIIString, extensionFa::ASCIIString, extensionTau::ASCIIString, initialQ::Float64, finalQ::Float64, dq::Float64, Np::Int64, r2dq::Float64, r2fa::Float64, scalesToRemove::Int64)

        #Load the data
        data = readdlm(inputfile,' ');
        x = data[:,1];
        y = data[:,2];

        Multifractal.ChhabraJensen(inputfile, extensionDq, extensionFa, extensionTau, x, y, initialQ, finalQ, dq, Np, r2dq, r2fa, scalesToRemove)

    end
    @time main("series.txt","tdq","tfa","tau",-5.0,5.0,1.0,9,-1.0,-1.0,1);

The first parameter of the function main is the input file with the time series you want to study the Multifractal spectrum. We will detail the other parameters later.

Multifractal.ChhabraJensen function returns as outputs 4 different files:

    series.tdq: 
    series.tfa:
    series.tau:
    summaryDq.dat: 

Later we will add details about the parameters given to the function as well as about the output files.