Simulate non-sparse data matrix X (and a corresponding normal y?)
simulate_regression_data(n, p)
n | number of samples |
---|---|
p | number of variables |
s | number of non-zero effects |
se | standard deviation of the residual |
pve | proportion of variance in Y explained by X |
misc | FILL IN OTHER PARAMS |
Y an n vector simulated gaussian (not centered or scaled)
Y_std an n vector simulated gaussian centered and scaled
beta a p vector of effects
sigma
sigma_std
mean_corX mean of correlations of effect variables (lower triangular entries of correlation matrix of effect variables)