Fig. 2From: Identifying biologically implausible values in big longitudinal data: an example applied to child growth data from the Brazilian food and nutrition surveillance systemTrajectories of length/height (A, B, C) and weight (D, E, F) of 1,000 randomly selected boys in the initial dataset (A and D) and after removing POs (B and E) and LOs (C and F). Notes: Longitudinal outliers (LOs) were based on the most conservative cutoff (<-3 / >+3). The black dashed line represents the observed mean curve based on all observations. To maintain scale and comparability there were observations that were not shown in panels A and D (available upon request)Back to article page