obs_h.npy:observation(56 variables)
obs_l.npy:observation(36 variables in truncation model)
true_h.npy:truth(56 variables)
true_l.npy:true(36 variables in truncation model)
sigma_hf.npy:standard deviation of high frequency(36 variables in truncation model)
sigma_hf_h.npy:standard deviation of high frequency(56 variables)
analysis.npz:DA results(['mua'])