null_dist: Generate null data with a specific distribution.

In nullabor: Tools for Graphical Inference

Description

Null hypothesis: variable has specified distribution

Usage

null_dist(var, dist, params = NULL)

Arguments

var	variable name
dist	distribution name. One of: beta, cauchy, chisq, exp, f, gamma, geom, lnorm, logis, nbinom, binom, norm, pois, t, unif, weibull
params	list of parameters of distribution. If NULL, will use fitdistr to estimate them.

Value

a function that given data generates a null data set. For use with lineup or rorschach

See Also

null_permute, null_lm

Examples

dframe <- data.frame(x = rnorm(150))
library(ggplot2)
# three histograms of normally distributed values
ggplot(
  data=rorschach(method=null_dist("x", "norm"), n = 3, true=dframe)
  ) +
  geom_histogram(aes(x=x, y=..density..), binwidth=0.25) +
  facet_grid(.~.sample) +
  geom_density(aes(x=x), colour="steelblue", size=1)

# uniform distributions are not as easy to recognize as such
dframe$x = runif(150)
ggplot(
  data=rorschach(method=null_dist("x", "uniform",
                 params=list(min=0, max=1)),
  n = 3, true=dframe)) +
  geom_histogram(aes(x=x, y=..density..), binwidth=0.1) +
  facet_grid(.~.sample) +
  geom_density(aes(x=x), colour="steelblue", size=1)

Example output

nullabor documentation built on Feb. 26, 2020, 1:07 a.m.