In alaindanet/ensupr:

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  eval = TRUE
)
library(datasets)

library(tidyverse)
library(magrittr)

data(iris)
head(iris)

# I do not like uppercase:
iris %<>% rename_all(tolower)
head(iris)

# I prefer tibble format:
iris %<>% as.tibble 
iris

We want to compute the mean value for each variable for each species. We will use the functions of the dplyr package.

iris_species <- iris %>%
  gather(variable, value, sepal.length:petal.width) %>%
  group_by(species, variable) %>%
  summarise(avg = mean(value)) %>%
  spread(variable, avg)

Tables

You can generate tables in markdown with kable() but you can also generate it in LaTeX.

knitr::kable(iris_species)

library(xtable)
xtable(iris_species)

Let's tweak a bit the table by removing row numbering, reduce digit number, etc... See r ?xtable and r ?print.xtable

iris_species_xtable <- xtable(iris_species, 
  caption = "Average flower phenology (in cm) of different species of the
  \textit{Iris} genus.",
  label = "tab:iris_species",
  align = c("c","l", "|", "r", "r", "r", "r"),
  digits = 1
  )
print(iris_species_xtable, include.rownames = FALSE)

Figures

iris_petall <- select(iris, species, petal.length)
p <- ggplot(iris_petall, aes(y = petal.length, x = species)) + 
  geom_violin()
p

This plot is not ready to be put in a paper (grey area, bad axes, font not ajusted). The package cowplot load a theme with better default. See the cowplot vignette.

library(cowplot)
p

You can also arrange plots in a grid:

p2 <- ggplot(mpg, aes(x = cty, y = hwy, colour = factor(cyl))) + 
  geom_point(size=2.5)
p2

plot_grid(p, p2, labels = "AUTO")

alaindanet/ensupr documentation built on May 18, 2019, 4:50 p.m.