generate_full_data: Generate full data

In timothy-barry/glmeiv: Implements the GLM-EIV model and method

Generate full data

Description

Generates B copies of a full GLM-EIV dataset.

Usage

generate_full_data(
  m_fam,
  m_intercept,
  m_perturbation,
  g_fam,
  g_intercept,
  g_perturbation,
  pi,
  n,
  B,
  covariate_matrix,
  m_covariate_coefs,
  g_covariate_coefs,
  m_offset,
  g_offset,
  run_mrna_unknown_theta_precomputation = FALSE,
  run_grna_unknown_theta_precomputation = FALSE,
  rm_covariate = "",
  mrna_duplet_rate = 0,
  grna_duplet_rate = 0
)

Arguments

m_fam	family for m
m_intercept	m intercept
m_perturbation	m perturbation coefficient
g_fam	family for g
g_intercept	g intercept
g_perturbation	g perturbation coefficient
pi	probability of perturbation
n	number of cells
B	number of i.i.d. datasets to generate
covariate_matrix	the (fixed) covariate matrix of confounding factors
m_covariate_coefs	coefficients for technical factors in model for m
g_covariate_coefs	coefficients for technical factors in model for g
m_offset	optional (fixed) offset vector for m
g_offset	optional (fixed) offset vector for g
run_unknown_theta_precomputation	optional; if TRUE, runs the negative binomial (unknown theta) precomputation on all datasets and stores the results ("m_precomp," "g_precomp") as attributes.

Value

a list of length B of synthetic datasets with columns (p, m, g)

Examples

## Not run: 
library(magrittr)
m_fam <- g_fam <- poisson() %>% augment_family_object()
pi <- 0.2; n <- 5000; B <- 5
m_intercept <- log(0.01); m_perturbation <- log(0.25); g_intercept <- log(0.005); g_perturbation <- log(2.5)
m_offset <- log(rpois(n = n, lambda = 10000)); g_offset <- log(rpois(n = n, lambda = 5000))
# no covariates
m_covariate_coefs <- g_covariate_coefs <- covariate_matrix <- NULL
dat_list <- generate_full_data(m_fam, m_intercept, m_perturbation, g_fam,
g_intercept, g_perturbation, pi, n, B, covariate_matrix, m_covariate_coefs,
g_covariate_coefs, m_offset, g_offset)
# with covariates
covariate_matrix <- data.frame(p_mito = runif(n, 0, 10))
m_covariate_coefs <- -0.1; g_covariate_coefs <- 0.2
dat_list <- generate_full_data(m_fam, m_intercept, m_perturbation, g_fam,
g_intercept, g_perturbation, pi, n, B, covariate_matrix, m_covariate_coefs,
g_covariate_coefs, m_offset, g_offset)

## End(Not run)

timothy-barry/glmeiv documentation built on Jan. 30, 2024, 3:46 p.m.