mkDf: Make a dataframe with simulated data given a set of...

Description Usage Arguments Details Value See Also Examples

Description

Generates data for a simulated experiment given population parameters and information about the design (one-factor design).

Usage

1
2
3
mkDf(nsubj = 24, nitem = 24, wsbi = FALSE,
  mcr.params = randParams(genParamRanges(), 1)[1, ],
  missMeth = "random", rigen = FALSE, verbose = FALSE)

Arguments

nsubj

number of subjects in the experiment

nitem

number of items in the experiment

wsbi

whether the design is between-items (TRUE) or within-items (FALSE)

mcr.params

population parameters (typically a row from the matrix generated by createParamMx

missMeth

method used for generating missing data (default is 'random'; alternatives are 'none', 'randomBig', 'bysubj', 'bycond', bysubjcond'; see Details)

rigen

use a 'random-intercepts-only' generative model (default FALSE)

verbose

Return all randomly generated values as well as the data.

Details

Note that for between-items designs, the variable w11 in the parameter vector is simply ignored. The default value for missMeth of 'random' will randomly remove between 0-5% of the observations. The option 'randomBig' will randomly remove 10-80%; 'bysubj' will randomly remove 10-80% of each subject's data; 'bycond' will randomly remove 10-80% of each condition's data; and 'bysubjcond' will randomly remove 10=80% of each subject/condition combination's data.

Value

If verbose = FALSE (the default), a dataframe, with fields:

SubjID

a factor, identifying subject number

ItemID

a factor, identifying item number

Cond

treatment condition, deviation coded (-.5, .5)

Resp

response variable

; If verbose = TRUE, a list, with elements:

dat

The data, with columns defined as above;

subj_re

by-subject random effects

item_re

by-item random effects

err

the residuals

See Also

genParamRanges, mkDf.facMixedAB

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
nmc <- 10
pmx <- cbind(randParams(genParamRanges(), nmc, 1001), seed=mkSeeds(nmc, 1001))

x.df <- mkDf(nsubj=24, nitem=24, mcr.params=pmx[1,], wsbi=FALSE)

## set rate of missing observations manually
pmx[,"miss"] <- 0.4
x.df2 <- mkDf(nsubj=24, nitem=24, mcr.params=pmx[1,], wsbi=FALSE)

## by-condition missing observation rates
pmx[,"pMin"] <- 0.1
pmx[,"pMax"] <- 0.8
x.df3 <- mkDf(nsubj=24, nitem=24, mcr.params=pmx[1,], wsbi=FALSE,missMeth="bycond")

## by-subject missing observation rates
x.df4 <- mkDf(nsubj=24, nitem=24, mcr.params=pmx[1,], wsbi=FALSE,missMeth="bysubj")

## by-subject/condition pair missing observation rates
x.df5 <- mkDf(nsubj=24, nitem=24, mcr.params=pmx[1,], wsbi=FALSE,missMeth="bysubjcond")

## verbose version
x.df <- mkDf(verbose = TRUE)
print(x.df$dat) # the data
print(x.df$subj_re) # by-subject random effects

dalejbarr/simgen documentation built on May 14, 2019, 3:32 p.m.