generate_meta_analysis: Generate meta-analysis data.
In replisims/peters-2016: Replication of Peters et al 2006

View source: R/data_generation_functions.R

Call all helper functions to generate a dataframe with the simulated data pertaining to a single meta-analysis.

generate_meta_analysis(
  job_id,
  scenario_id,
  bias_type,
  bias_percentage = NULL,
  bias_strength = NULL,
  odds_ratio,
  heterogeneity,
  ma_size,
  prob_cg_distr,
  n_cg_distr,
  bias_table
)

`job_id`	User defined job_id. Used for determining a unique seed.
`scenario_id`	Unique id fer scenario = constellation of simulation parameters.
`bias_type`	Bias type can either be "p" or "es".
`bias_percentage`	Percentage of studies that will be removed to to publication bias. Only needs to be provided when bias type = "es".
`bias_strength`	String indicating bias severity "moderate" or "strong". only needs to be supplied when bias type = "p".
`odds_ratio`	True underlying effect of the symulated meta-analysis.
`heterogeneity`	Numeric value indicating the heterogeneity fraction.
`ma_size`	Intended number of studies to be included in final meta-analysis (after publication bias).
`prob_cg_distr`	Probablility of an event in the control group. Can be any value or function that evaluates to a numeric value between 0 and 1.
`n_cg_distr`	Sample size in control group (= sample size in exposed group) Can be any value or function that evaluates to an integer.
`bias_table`	List of pvalue cut-offs and corresponding selection probabilities list element names must correspond to elements of bias strength #' @return A dataframe with ma-size rows. A subset of rows returned by simulate_unbiased_study_set()