Normalizing and exporting impulse responses

Lukas Freund

2017 8 4

0 个回答

11 次查看（30 天）

0 个投票

sample.mat

Hi all,

I am a relatively new matlab user facing the following problem which I hope you can help me with.

Given: structural impulse responses (SIRs) based on a simple VAR; structure has three elements (irfinf, irfmed, irfsup). Saved in 'filenameofIR.mat'
Goal(s)

1. Export data to excel such that I have all three elements (median, upper and lower confidence bound) in one worksheet.

2. Normalize impulse responses For instance, I have an exogenous variable as the first variable in the VAR and government spending as the second variable. I wish to normalize the impulse responses such that the max impact of the first variable on government spending is 1 (to facilitate comparison...).

I attach a sample of the sort of VAR output I am taking as my starting point.

Approach so far For 1.

 data=load('filenameofIR.mat');
fn=fieldnames(data); %get all variable names
IR=data.(fn{1});  %get the first element of the structure, this is irfinf
xlswrite('desiredxlsfilenname',IR1(:,:,1)) %export e.g. the first impulse response (i.e. with the exogenous variable as impulse); but this is only the median, I would like to have all three together.

For 2., I have not found any useful guidance at all /:

I would greatly appreciate your help!

Best,

Lukas

4 个评论
显示 2更早的评论隐藏 2更早的评论

Lukas Freund 2017-8-17

Thank you, José-Luis. However, given this normalization, surely the entire shape of the impulse response is distorted, as I only want to normalize the functions relative to one max...