27pKH-8 競馬単勝オッズの時系列データ解析とハードの法則(経済物理学2,領域11(統計力学,物性基礎論,応用数学,力学,流体物理))

Abstract

A method is proposed for analysing daily mood scores for premenstrual tension (PMT) in single menstrual cycles by fitting an approximation function consisting of the first two harmonics of a Fourier series. On the basis of harmonic analysis of 133 cycles of daily mood scores obtained from women perceiving themselves to have the premenstrual syndrome (PMS), the severity of PMT is taken to be the difference between the mean and the fitted function at a point 92% of the way through the menstrual cycle. The significance of the severity is calculated statistically. Significant PMT was detected in 70% of the 133 cycles; this compared with an 8% incidence in women who considered that they did not have PMS. The method improves on previously described methods for detecting PMT by: (1) using statistical criteria; (2) modelling cycles with asymmetric and double peaks; (3) using a difference rather than a ratio as the severity estimate; (4) smoothing the data and (5) giving valid estimates when data points are missing. There was no evidence for the presence of distinct periodicities with wavelengths less than half that of the menstrual cycle.