This block is a continuation of a previous one, and an experimentation of how to rebuilt an approximation of a timeline thanks to an inverse Fast Fourier Transform (iFFT) algorithm.
Usages :
- in the top graph, Drag & Drop points to update the timeline, or use the shortcuts below the graph; this will update the corresponding FFT components (in the bottom graph)
- in the bottom graph, move the treshold up and down; this will rebuild another timeline (in the top graph) based on the retained FFT components
Notes:
I’ve done other bl.ocks dealing with timeline analyses:
- another block highlights how important detrending is when trying to detect seasonality
- another block experiments seasonality detection
- another block experiments autocorrelation
- another block experiments time series correlation
- another block deals with the impact of seasonality when computing the trend of a timeline
Acknowledgments: