Data and software underlying the publication: Waiting for Forcot: Accelerating FORC Processing 100x using a Fast-Fourier-Transform Algorithm
datasetposted on 24.10.2019 by Thomas Berndt, L. (Liao) Chang
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
First-order reversal-curves (FORC) are a widely used tool to analyze magnetic mineralogy and domain states, but require extensive processing -- in particular smoothing -- to be plotted as FORC diagrams. Currently, smoothing is a computationally complex task involving repeated least-squares surface optimization routines, sometimes taking minutes to compute for high-resolution FORCs (leaving users with the feeling of ``waiting for Godot'', who never comes). Here we show how the same computation can be carried out much more efficiently in Fourier space and present a new FORC processing software, called Forcot. The new algorithm, combined with an improved user-interface enables users to create print-quality FORC diagrams within a few seconds. Processing times are shown to be reduced by factors from 2 to 100 depending on size and smoothing factor (SF) compared to existing FORC smoothing algorithms. Additionally, optimal SF can be determined directly from the noise spectrum in Fourier space and does not require repeated smoothing of diagrams as in previous programs. Finally, formatting of figures is done automatically by our software such that diagrams can be directly used for print.