Application and interpretation of functional data analysis techniques to differential scanning calorimetry data from lupus patients
Fig 3
Smoothed regression coefficients estimated for predicting thermograms from disease status, sex, race, and year of birth.
From left to right, top to bottom, panel 1 is the intercept coefficient, corresponding to the overall mean thermogram. The second panel shows the estimated perturbation (regression coefficients) of the overall mean needed to fit a curve for cases. The third panel shows the estimated perturbation of the overall mean needed to fit a curve for females. Panels 4–6 show the estimated perturbation of the overall mean thermogram needed to fit a curve for individuals with a birth year in (1944, 1995], (1955, 1971], and (1971, 1993], respectively. Finally, panel 7 shows the estimated perturbation of the overall mean thermogram needed to fit a curve for individuals identifying as White.