How to do Gaussian Filter 1D?

Hi, Community

I wanna ask about how to do a Gaussian Filter in just 1D data.... So i have a data vector based on time series like this :

And i want to filter the data by using Gaussian Filter. So, Anyone, How to do that? Thank you so much....

Answer 1

[Chunru]

 x = readtimetable("IAGA Daily Magnetic Data (1m) Extraction 04-Jul-2021 (CTS).txt")

x = 1440×8 timetable
       Time         Var1      Var2     Var3     Var4     Var5    Var6     Var7     Var8 
    __________    ________    _____    _____    _____    ____    _____    _____    _____

    2021-07-04    00:00:00    22494    47892    47882     0      22494    23.58    42280
    2021-07-04    00:01:00    22494    47891    47882     0      22494    23.54    42280
    2021-07-04    00:02:00    22494    47891    47882     0      22494    23.58    42280
    2021-07-04    00:03:00    22494    47891    47882     0      22494    23.51    42280
    2021-07-04    00:04:00    22494    47891    47882     0      22494    23.62    42280
    2021-07-04    00:05:00    22494    47891    47882     0      22494    23.77    42280
    2021-07-04    00:06:00    22494    47891    47881     0      22494    23.84    42280
    2021-07-04    00:07:00    22494    47891    47881     0      22494    23.88    42280
    2021-07-04    00:08:00    22494    47891    47881     0      22494    23.92    42280
    2021-07-04    00:09:00    22494    47891    47882     0      22494    23.99    42280
    2021-07-04    00:10:00    22494    47891    47882     0      22494    24.07    42280
    2021-07-04    00:11:00    22494    47892    47882     0      22494    24.18    42280
    2021-07-04    00:12:00    22494    47892    47882     0      22494    24.18    42280
    2021-07-04    00:13:00    22495    47892    47882     0      22495    24.33    42280
    2021-07-04    00:14:00    22495    47892    47882     0      22495    24.37    42280
    2021-07-04    00:15:00    22495    47892    47882     0      22495    24.33    42280

 w = gausswin(10, 2.5);
 w = w/sum(w);
 y = filter(w, 1, x.Var7);
 plot(x.Var7, 'r'); hold on; plot(y, 'b');
 legend('original', 'filtered')

[Tyann Hardyn]

Thats so Great, Sir. I ll try with Gausswin function right now. Because Iam trying by using :

data_noise = fullfile("Path", "IAGA Daily Magnetic Data (1m) Extraction 04-Jul-2021 (CTS).txt")
data_noise0 = detectImportOptions(data_noise, 'NumHeaderLines',0,'Delimiter', ' ', 'Whitespace', ' ',...
'ConsecutiveDelimitersRule', 'join', 'EmptyLineRule', 'skip', 'FileType','text');

data0_komph = data0.Var3;
time_ori = string(data0.Var2);
l_wak = length(data0_komph);
gauss_filt = fspecial('gaussian',[l_wak 1] ,10);
gauss_komph = conv(data0_komph,gauss_filt, 'same');

plot(time_ori, gauss_komph);

and the result is weird :

Its maybe the both of side (gauss filter) part which is symmetry was still remain....

Iam so grateful that someone can give me an Input Idea.... Thank You So Much

[Tyann Hardyn]

Hi, Chunru

I have a problem when plotting the gaussian filter result and the original data....

What is x component you re used in Plot? Why it become a range between 0 - 1500 instead of time series data, which mean from 2021-07-04 00:00:00 until 2021-07-04 23:59:00?

I actually have a problem in Var6 Plot. I need to filter the var6 (Horizontal Magnetic Component) and not in var7