A geom to fit a LM model to a short time series

Usage

geom_lm(
  mapping = NULL,
  data = NULL,
  stat = "LM",
  position = "identity",
  show.legend = NA,
  na.rm = FALSE,
  inherit.aes = TRUE,
  warn = TRUE,
  n = 10,
  nBootSamples = 499,
  pValThreshold = 0.05,
  ...
)

Arguments

mapping

Set of aesthetic mappings created by aes(). By default inherit.aes = TRUE, which assigns the top-level plotting aes() to the GLS geom.

data

Input series to be analyzed. If NULL, data is inherited from previous layer or ggplot call.

stat

stat

position

position

show.legend

show legend

na.rm

remove NAs

inherit.aes

inherit aesthetics

warn

Conditional. If TRUE, a warning message will be returned when N < 30.

n

Numeric. Number of points to use for trend. Default = 10.

nBootSamples

Numeric. Number of bootstrap samples used to test Null hypothesis. Default = 499

pValThreshold

Numeric. Significance level of the test. Default = 0.05

...

Other arguments may be passed to the stat, including fixed aesthetics.

Value

If slope is significantly different from 0 (p < 0.05), then a line of best fit is returned. Otherwise output is NULL.

Examples

library(ggplot2)

#Generate series

m <- 0.1
x <- 1:10
y <-  m*x + rnorm(10, sd = 0.35)

data <- data.frame(x = x,
                  y = y)

#Plot series with trend
ggplot(data = data) +
  geom_line(aes(x = x, y = y)) +
  geom_lm(aes(x = x, y = y), n = 10)