generate.sim.parameters.dt: Generates parameters for simulating data for the study e.g....

Description Usage Arguments Value

View source: R/simulate.helpers.R

Description

Generates parameters for simulating data for the study e.g. what is the recruitment rate, mean, intervention effect etc in each cluster.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
generate.sim.parameters.dt(
  cluster.dt,
  sequence.dt,
  ppt.per.unit.time.mean = 1,
  target.n.participants = NA,
  ppt.per.unit.time.sd = 0.1,
  ppt.per.unit.time.sd.proportion.of.mean = NA,
  base.outcome = 10,
  intervention.effect.mean = 1,
  intervention.effect.sd = 0.1,
  intervention.effect.sd.proportion.of.mean = NA,
  cluster.effect.mean = 0,
  cluster.effect.sd = 0.1,
  cluster.effect.sd.proportion.of.mean = NA,
  time.effect.per.unit.mean = 0.01,
  time.effect.per.unit.sd = 0.001,
  time.effect.per.unit.sd.proportion.of.mean = NA,
  individual.noise.mean = 0.2,
  individual.noise.sd = 0.04,
  individual.noise.sd.proportion.of.mean = NA,
  cycle.amplitude.mean = 0.8,
  cycle.amplitude.sd = 0.1,
  cycle.amplitude.sd.proportion.of.mean = NA,
  cycle.start.mean = 10,
  cycle.start.sd = 0,
  cycle.start.sd.proportion.of.mean = NA,
  cycle.period.mean = 20,
  cycle.period.sd = 0,
  cycle.period.sd.proportion.of.mean = NA,
  min.time = 0,
  max.time = NA,
  n.to.generate = 1
)

Arguments

cluster.dt

A study info data.table, containing data about clusters.

sequence.dt

A study info data.table, containing data about sequences.

ppt.per.unit.time.mean

What is the mean recruitment rate across clusters per one time.

target.n.participants

Integer how many participants, will adjust ppt.per.unit.time.mean in sim.parameters.dt by weighting them to get the desired number (Default: NA)

ppt.per.unit.time.sd

What is the variance in the recruitment rate between clusters.

ppt.per.unit.time.sd.proportion.of.mean

What is the variance in the recruitment rate between clusters as a multiple of mean. Will override ppt.per.unit.time.sd, disabled by default.

base.outcome

The outcome at every cluster before adding any effects

intervention.effect.mean

The mean effect of the intervention across cluster.

intervention.effect.sd

The variance in the effect of the intervention between clusters.

intervention.effect.sd.proportion.of.mean

The variance in the effect of the intervention between clusters as a multiple of mean. Will override intervention.effect.sd, disabled by default.

cluster.effect.mean

The mean effect associated with each cluster.

cluster.effect.sd

The variance in the effect associated with cluster.

cluster.effect.sd.proportion.of.mean

The variance in the effect associated with cluster as a multiple of mean. Will override cluster.effect.sd, disabled by default.

time.effect.per.unit.mean

The mean effect of time per one unit across clusters.

time.effect.per.unit.sd

The variance in the effect of time per one unit between clusters.

time.effect.per.unit.sd.proportion.of.mean

The variance in the effect of time per one unit between clusters as a multiple of mean. Will override time.effect.per.unit.sd, disabled by default.

individual.noise.mean

The mean of the standard deviation used to generate individual noise across clusters (i.e. generated by rnorm).

individual.noise.sd

The variance in the standard deviation used to generate individual noise across clusters.

individual.noise.sd.proportion.of.mean

The variance in the standard deviation used to generate individual noise across clusters as a multiple of mean. Will override individual.noise.sd, disabled by default.

cycle.amplitude.mean

The mean amplitude of the cyclical effect across clusters.

cycle.amplitude.sd

The variance in the amplitude of the cyclical effect between clusters.

cycle.amplitude.sd.proportion.of.mean

The variance in the cycle start time between clusters as a multiple of mean. Will override cycle.start.sd, disabled by default.

cycle.start.mean

The mean cycle start time across clusters.

cycle.start.sd

The variance in the cycle start time between clusters.

cycle.period.mean

The mean cycle period across clusters.

cycle.period.sd

The variance in the cycle period between clusters.

cycle.period.sd.proportion.of.mean

The variance in the cycle period between clusters as a multiple of mean. Will override cycle.period.sd, disabled by default.

min.time

Minimum time value (Default: 0)

max.time

Maximum time value, if NA will be calculated from sequence.dt, by assuming equal period lengths, which may not be correct (Default: NA)

n.to.generate

Integer how many different effects to generate. If more than one, an integer column will be added (sim.number) to delineate. (Default: 1)

Value

The provided cluster.dt parameter, with simulation parameters added as columns.


mattmoo/SteppedWedgeAnalysis documentation built on Jan. 14, 2020, 12:25 a.m.