Skip to contents

Preprocessing, description and figures for clinical cohort data stored in wide format: one row per subject, one column per variable and visit, with column names following the <ROOT>_<VISIT> convention (e.g. NFL_V0, NFL_V3).

Details

Functions are grouped in families sharing a prefix:

Limits of detection and cut-offs are handled by the companion package lodestar.

Symbols in figures and PDF output

Figures use Unicode symbols (Greek letters such as rho and Delta, minus and comparison signs, superscript exponents). The classic grDevices::pdf() and grDevices::postscript() devices cannot draw characters outside Latin-1; this is the device of R CMD check and of ggplot2::ggsave() for a .pdf file. The option vernissage.unicode controls the symbols:

  • "auto" (default): Unicode, except when the open device (or the default device, when none is open) is a classic pdf() or postscript() device; then ASCII equivalents are used (rho, Delta, >=, x10^-4).

  • TRUE / FALSE: always Unicode / always ASCII.

The choice is made when the figure is built. To keep the symbols in a PDF file, save with a Cairo device, e.g. ggplot2::ggsave("figure.pdf", p, device = grDevices::cairo_pdf); to build figures in an interactive session and save them later with pdf(), set options(vernissage.unicode = FALSE).

Author

Maintainer: Clément Poupelin clementjc.poupelin@gmail.com

Authors: