In ymer/jsmp: jsmp

knitr::opts_chunk$set(
  echo = F, message = F, warning = F,
  fig.path = "man/figures/",
  collapse = T,
  comment = "#>"
  )

library(jsmp)

jsmp

This package does variety of things I find useful when working with R Tidyverse:

• Import commonly used packages
• Changes the default ggplot theme
• Overrides dplyr::summarise to remove leftover grouping
• Contains various functions for use with ggplot
• Contains various functions for use in tidy code

Installation

devtools::install_github("ymer/jsmp")

Custom ggplot theme

mtcars |> ggplot(aes(x = mpg, y = disp)) +
   geom_point()

ggplot functions

• gg_hist_percent : Draws a histogram with y-axis as a percentage

• gg_legend_remove : Removes the legend

• gg_legend_notitle : Removes the legend title

• gg_legend_bottom : Moves the legend to the bottom of the plot

• gg_loess_line : Plots a locally weighed regression line

• gg_regression_line : Plots a regression line

• gg_x_remove : Removes the x-axis

• gg_x_rotate : Rotates x-axis labels 45°

• gg_x_wrap : Wraps text on x-axis labels

• gg_y_big : Changes the y-axis to avoid scientific notation for big numbers

• gg_y_percent : Changes the y-axis to percentage

• gg_y_percent_zero : Changes the y-axis to percentage, and adjusts the y-axis to start at 0.

• gg_y_remove : Removes the y-axis

• gg_y_zero : Adjusts the y-axis to start at exactly 0

General functions

• summarise : Overrides the standard summarise function, so that leftover grouping is dropped after summarising

mtcars |> 
   group_by(am, gear) |> 
   dplyr::summarise(mean_mpg = mean(mpg)) |> 
   group_vars()

mtcars |> 
   group_by(am, gear) |> 
   summarise(mean_mpg = mean(mpg)) |> 
   group_vars()

• %notin% : As %in% but exclusionary

• d : Formats table (using gt as base)

• ci_means : Finds the means and confidence intervals

ToothGrowth |> 
   group_by(supp) |> 
   ci_means(len) |> 
   ggplot(aes(y = supp, x = mean)) +
   geom_crossbar(aes(xmin = ci.lower, xmax = ci.upper), width = 0.3, size = 0.7, color = c1)

• ci_proportions : Finds the proportions and confidence intervals

mtcars |> 
   mutate(cyl = as.character(cyl)) |> 
   count(cyl) |> 
   mutate(total = sum(n)) |> 
   ci_proportions() |> 
   ggplot(aes(x = cyl, y = proportion)) +
   geom_col(position = position_dodge()) +
   gg_y_percent_zero() +
   gg_legend_notitle() +
   labs(y = "", x = "") +
   geom_errorbar(aes(ymin = ci.lower, ymax = ci.upper),
                  width=.2, position=position_dodge(.9))

• do_if : Use a condition in a pipe flow

only_high_values <- T
mtcars |> do_if(only_high_values, ~ . |> filter(disp > 180)) |> 
   summarise(mean(disp))

only_high_values <- F
mtcars |> do_if(only_high_values, ~ . |> filter(disp > 180)) |> 
   summarise(mean(disp))

high_or_low <- "low"
mtcars |> do_if(high_or_low == "high", 
                 ~ .x |> filter(disp >= 180), 
                 ~ .x |> filter(disp < 180)) |> 
   summarise(mean(disp))

• filter_duplicates : selects duplicated rows

mtcars |> filter_duplicates(wt)

• fix_names : Changes the column names to tidy style

iris |> fix_names() |> 
   head()

• left_join0 : Performs a left_join, while setting values in missing rows to 0 instead of NA.

df1 <- tribble(
   ~id, ~v1,
   1, 2,
   2, 2,
   3, 10)

df2 <- tribble(
   ~id, ~v2,
   1, 2,
   3, 4)

left_join(df1, df2)

left_join0(df1, df2)

• percent : Returns proportion formatted as percentage

percent(0.173234235)

• rows : Facilitates looping

df <- mtcars |> head()
for (row in rows(df)){
   print(row$mpg)
}

• tab : Ordered count including percentage

ToothGrowth |>
   filter(len > 20) |> 
   tab(supp)

• transpose : Flips rows and columns

mtcars |> 
   rownames_to_column() |> 
   transpose()

Imported packages

• broom
• glue
• gt
• lubridate
• magrittr
• patchwork
• stringr
• tidyverse

ymer/jsmp documentation built on Jan. 2, 2022, 6:17 p.m.