Assigning features to the matrix based on windowing
7 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Hello,
At the beginning I have at my disposal file with 3 columns (signal sample, condition, device):
-297.367190000000 8 1
-295.132890000000 8 1
-282 8 1
-268.007000000000 8 1
-262.109380000000 8 1
-263.296880000000 8 1
-263.562500000000 8 1
-258.973210000000 8 1
-255.209820000000 8 1
-252.130060000000 8 1
-247.141450000000 8 1
-244 8 1
-243.025000000000 8 1
-237.704170000000 8 1
-230.770830000000 8 1
-230.597220000000 8 1
-239.697970000000 8 1
-254.455570000000 8 1
-268.051250000000 8 1
-269.360000000000 8 1
-255.611150000000 8 1
100.781000000000 3 2
54.1178190000000 3 2
25.2507100000000 3 2
23.8387100000000 3 2
46.2026210000000 3 2
49.5785290000000 3 2
20.7356310000000 3 2
-16 3 2
-73.7242630000000 3 2
Then using a the presented code for windowing I extract the following features:
fs = 256;
sbin=4;
%sbin = 1; %1s
window=fs*sbin;
overlap=fs;
Nwin = floor((length(data)-window)/overlap)+1;
for k=1:Nwin
feature{1,1}(k) = var(data(1,(k-1)*overlap+1:(k-1)*overlap+window));
[Higuchi_FD(k)] = feature2(data(1,(k-1)*overlap+1:(k-1)*overlap+window),window);
feature{2,1}(k) = feature2(k);
[Katz_FD(k)] = feature3(data(1,(k-1)*overlap+1:(k-1)*overlap+window));
feature{3,1}(k) = feature3(k);
for instance it gives the following results (the information is included in the rows):
{[ 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.01 … ]}
{[ 1.21 1.21 1.20 1.21 1.22 1.21 1.21 1.21 1.21 1.20 1.21 1.21 1.20 1.21 1.21 1.20 1.19 … ]}
{[ 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 … ]}
At the end I need to assign the information about condition and device to the windowed features. There 9 condition and 3 devices. How to perform that?
Regards
Elzbieta
1 commentaire
Voss
le 6 Sep 2024
Please share all code and data necessary for us to run the code you've shown here. You can upload files using the paperclip button.
Réponses (2)
Shishir Reddy
le 6 Sep 2024
Hey Elzbieta
Assigning the information about condition and device to the windowed features can be done by determining the most frequent condition and device for each window and assigning them to the features. This can be implemented as shown in the following code snippet -
for k = 1:Nwin
start_idx = (k-1) * overlap + 1;
end_idx = (k-1) * overlap + window;
windowed_data = signal_samples(start_idx:end_idx);
windowed_conditions = conditions(start_idx:end_idx);
windowed_devices = devices(start_idx:end_idx);
features{1, 1}(k) = var(windowed_data);
features{2, 1}(k) = feature2(windowed_data, window); % Assuming feature2 is defined
features{3, 1}(k) = feature3(windowed_data); % Assuming feature3 is defined
conditions_windowed(k) = mode(windowed_conditions); % or conditions(start_idx)
devices_windowed(k) = mode(windowed_devices); % or devices(start_idx)
end
Within each window, the condition and device are determined. As conditions or devices may vary within the window, the most frequent value (‘mode’ in MATLAB) is taken.
This approach ensures that each set of features extracted from a window is accurately labelled with the appropriate condition and device.
I hope this helps.
Image Analyst
le 6 Sep 2024
Why not just tack on the condition and device as the last two rows in your loop?
feature{4}(k) = data(k, 2);
feature{5}(k) = data(k, 3);
0 commentaires
Voir également
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!