Issue with one-sided HP filter

Estimation ML/Bayesian estimation
1

Dear all,

I am new to Bayesian estimation and was trying to implement Bayesian estimation. I have collected data, applied log-transformation, but when I use one-sided HP filter from https://git.dynare.org/Dynare/dynare/-/blob/master/matlab/one_sided_hp_filter.m, I see almost no difference between the data before and after filter application, however I am expecting to arrive at fluctuations around the zero. It seems that I miss something. I have read guide by Professor Pfeifer, but still puzzled with the result. For example, I try to detrend consumption (C), but results are strange. Please see the data I use in the file attached.
final_data.mat (3.8 KB)

2

The issue was that I applied filter to the data directly via command window and did not store output. To resolve this, you have to write a separate .mat script, which would call the desired HP-filter function and store the output, i.e. [ytrend,ycycle], as a separate variable. Afterwards, you use the output to rewrite the desired variable.