Simulated trajectories from a process convolution model for movement conditioned on a latent social network. The inset in the bottom right shows the dynamic pairwise relationships among the three individuals. A time interval during which individuals 1 and 3 are strongly connected is highlighted with a gray box. On the trajectories, black arrows oriented in the direction of movement, with arrowhead length proportional to the connection strength, highlight the impact of this particular connection on movement. At the onset of the strong connection period, the indicated paths orient toward each other as individuals demonstrate attractive behavior. As the connection strength weakens, the individuals resume independent movement.

Graphic illustrating the questions leading to the choice of different type of multiple testing methods in integrative genomics.

Covariance for longitudinal data can be modeled through generalized auto‐regressive coefficients and innovation variances. Using the functional ANOVA model framework, we propose to model the varying coefficient function for dependence among ordered measurements with main effects of the time lag and the orthogonal additive direction, and their interaction.