stepwise_coco: stepwise_conditional

In theboocock/coco: Correlation-based conditional analysis

Description

Stepwise coco parameter estimation wrapper.

Forward model selection in coco.

Usage

stepwise_coco(res_preparation, p_value_threshold = 1e-04,
  colinear_threshold = 0.9, joint = T, max_iter = 100,
  return_all_betas = F)

stepwise_coco(res_preparation, p_value_threshold = 1e-04,
  colinear_threshold = 0.9, joint = T, max_iter = 100,
  return_all_betas = F)

Arguments

p_value_threshold	float P-value threshold for forward-stepwise selection.
joint	boolean (T) = perform joint estimation when doing stepwise selection. (F) = perform conditional estimation when doing stepwise selection. @export
data_set	data.frame. Coco formatted input data.
ld_matrix	matrix. LD matrix ordered by postiion
rsids	character vector. SNPIDs in res preparation to perform CoCo analysis on.
idxs	integer vector - postiion in input data to condition on.
var_y	float. Phenotypic variance for Y, if not specified it will be estimated using estimate_vary.
hwe_variance	boolean. Force the use of Hardy-weinberg equillibrium when calculating the genotypic variance.
data_set	input file format
ld_matrix	LD matrix for the subset of the region.
stepwise_results	- results from a stepwise analysis.

Details

stepwise_coco wraps joint_from_ids for convinience.

stepwise_coco. Forward model selection in coco.

Value

all_but_ones betas and standard error for the entire region.

Author(s)

James Boocock and Eli Stahl

James Boocock

theboocock/coco documentation built on May 31, 2019, 9:10 a.m.