statar (version 0.2)

graph: Experimental function to graph a dataset

Description

Experimental function to graph a dataset

Usage

graph(x, ..., along_with = NULL, by = NULL, w = NULL, reorder = TRUE,
  winsorize = TRUE, facet = FALSE, verbose = FALSE, type = if
  (is.null(substitute(along_with))) {     "density" } else {     if
  (is.null(formula))          "loess"     else "felm" }, formula = NULL,
  n = 20)
graph_(x, ..., .dots, along_with = NULL, by = NULL, w = NULL,
  reorder = TRUE, winsorize = TRUE, facet = FALSE, verbose = FALSE,
  type = if (is.null(along_with)) {     "density" } else {     if
  (is.null(formula))          "loess"     else "felm" }, formula = NULL,
  n = 20)

Arguments

x

A data.table.

...

Variables to include. Defaults to all non-grouping variables. See the select documentation.

along_with

A variable that specifies the x axis. Should be specified when "type" is "line", "lm" or "loeless".

by

Groups within which variables should be ploted.

w

Analytical weights (experimental)

reorder

Should the category with the most count be printed first?

winsorize

Should variables winsorized?

facet

Should different groups graphed in different windows?

verbose

Should warnings (regarding missing values, outliers, etc) be printed?

type

type of graph among "density", "boxplot", "line", or a statistical method (like "lm", "loeless" or "felm")

formula

to use then type is "felm"

n

Number of quantiles to plot

.dots

Used to work around non-standard evaluation.

Examples

library(data.table)
N <- 1e2
DT <- data.table(
  id = sample(c("id1","id2","id3"), N, TRUE),
  v1 = sample(c(1:5), N, TRUE),
  v2 = rnorm(N, sd = 20),
  v3 = sample(runif(100, max=100), N, TRUE)
)
DT[, v4 := v3 + rnorm(N, sd = 20)]
graph(DT)
graph(DT, by = id)
graph(DT, by = id, facet = TRUE)
graph(DT, by = id, type = "boxplot")
graph(DT, v3, along_with = v2, by = id, type = "loess")