Autocorrelation, Enright and DFT method's performance in recovering the fast and the slow period of <i>in silico</i> bi-oscillatory data as a function of the number of points per cycle (<i>NPC</i>), with 0% noise.

<p>(A) Both Enright and DFT method's relative differences (<i>RD</i>) remain fairly constant on average with increasing <i>NPC</i> while recovering the fast simulated period. The Enright method returns on average, better results than the Autocorrelation or DFT methods. The Autocorrelation method is the least accurate, producing average differences starting around 60% and decreasing to 35% with 20 <i>NPC</i>. (B) The DFT method produces the most accurate results, but overall, average differences are approximately 30%. The red diamond, the green square and the blue circle represent the Autocorrelation, Enright and DFT methods, respectively.