control_chart: Create a control chart

In healthcareai: Tools for Healthcare Machine Learning

Create a control chart

Description

Create a control chart, aka Shewhart chart: https://en.wikipedia.org/wiki/Control_chart.

Usage

control_chart(
  d,
  measure,
  x,
  group1,
  group2,
  center_line = mean,
  sigmas = 3,
  title = NULL,
  catpion = NULL,
  font_size = 11,
  print = TRUE
)

Arguments

d	data frame or a path to a csv file that will be read in
measure	variable of interest mapped to y-axis (quoted, ie as a string)
x	variable to go on the x-axis, often a time variable. If unspecified row indices will be used (quoted)
group1	Optional grouping variable to be panelled horizontally (quoted)
group2	Optional grouping variable to be panelled vertically (quoted)
center_line	Function used to calculate central tendency. Defaults to mean
sigmas	Number of standard deviations above and below the central tendency to call a point influenced by "special cause variation." Defaults to 3
title	Title in upper-left
catpion	Caption in lower-right
font_size	Base font size; text elements will be scaled to this
print	Print the plot? Default = TRUE. Set to FALSE if you want to assign the plot to a variable for further modification, as in the last example.

Value

Generally called for the side effect of printing the control chart. Invisibly, returns a ggplot object for further customization.

Examples

d <-
  tibble::tibble(
    day = sample(c("Monday", "Tuesday", "Wednesday", "Thursday", "Friday"),
                 100, TRUE),
    person = sample(c("Tom", "Jane", "Alex"), 100, TRUE),
    count = rbinom(100, 20, ifelse(day == "Friday", .5, .2)),
    date = Sys.Date() - sample.int(100))

# Minimal arguments are the data and the column to put on the y-axis.
# If x is not provided, observations will be plotted in order of the rows

control_chart(d, "count")

# Specify categorical variables for group1 and/or group2 to get a separate
# panel for each category

control_chart(d, "count", group1 = "day", group2 = "person")

# In addition to printing or writing the plot to file, control_chart
# returns the plot as a ggplot2 obejct, which you can then further customize

library(ggplot2)
my_chart <- control_chart(d, "count", "date")
my_chart +
  ylab("Number of Adverse Events") +
  scale_x_date(name = "Week of ... ", date_breaks = "week") +
  theme(axis.text.x = element_text(angle = -90, vjust = 0.5, hjust=1))

healthcareai documentation built on Sept. 5, 2022, 5:12 p.m.