statsmodels.stats.meta_analysis._fit_tau_iterative

statsmodels.stats.meta_analysis._fit_tau_iterative(eff, var_eff, tau2_start=0, atol=1e-05, maxiter=50)[source]

Paule-Mandel iterative estimate of between random effect variance

implementation follows DerSimonian and Kacker 2007 Appendix 8 see also Kacker 2004

Parameters:

  • eff (ndarray) – effect sizes

  • var_eff (ndarray) – variance of effect sizes

  • tau2_start (float) – starting value for iteration

  • atol (float, default: 1e-5) – convergence tolerance for absolute value of estimating equation

  • maxiter (int) – maximum number of iterations

Returns:

  • tau2 (float) – estimate of random effects variance tau squared

  • converged (bool) – True if iteration has converged.