estimation.effective_count_matrix is slow for large data sets, especially with the setting truncate_acf=False. This can probably be sped up by a lot.
Essentially a lot of damped autocorrelation functions are computed which itself could be sped up. In addition the looping through the acf's from different trajectories is probably unnecessarily slow.
estimation.effective_count_matrix is slow for large data sets, especially with the setting truncate_acf=False. This can probably be sped up by a lot. Essentially a lot of damped autocorrelation functions are computed which itself could be sped up. In addition the looping through the acf's from different trajectories is probably unnecessarily slow.