# How to up-sample both signals and labels together?

3 ビュー (過去 30 日間)
Mibang 2024 年 3 月 14 日
コメント済み: Mibang 2024 年 3 月 15 日
I have ECG-signals and corresponding labels with the sampling rate of 250 Hz (attached).
Attached "sig.mat" is the ECG-signal and "labl.mat" is the labels, where five signals and five labels (5x1 cell).
I want to up-sample these signals and labels coherently (synchronized in time after upsampling) with 360 Hz.
I heard I can use "resample" function for the signal, but I don't know it for the labels.
I may use labeler app, but this is too slow.
I would appreciate it if anyone could help me on how to label automatically in accordance with the up-sampled signals.
Thanks,

サインインしてコメントする。

### 採用された回答

Chunru 2024 年 3 月 15 日

fs = 250;
fsnew = 360;
for i =1:length(sig)
% signal and label
s = sig{i};
l = labl{i};
% categories
c = categories(l);
% original and new time
t = (0:length(s)-1)/fs;
tnew = t(1):(1/fsnew):t(end);
% interpolation
snew = interp1(t, s, tnew, 'linear');
lnew = interp1(t, double(l), tnew, 'nearest');
% convert back to categorical data
lnew = categorical(lnew, 1:length(c), c);
signew{i, 1} = snew;
lablnew{i, 1} = lnew;
end
sig
sig = 5×1 cell array
{[ -0.4456 -0.3906 -0.4456 -0.5006 -0.4731 -0.5555 -0.5830 -0.6380 -0.5280 -0.5280 -0.6105 -0.6380 -0.5280 -0.6105 -0.5555 -0.5830 -0.5280 -0.5280 -0.5830 -0.5555 -0.4731 … ] (1×5000 double)} {[-0.0973 -0.0211 0.0297 -0.0211 -0.0211 0.0805 0.1567 0.1821 0.1821 0.2329 0.3345 0.3345 0.3599 0.4107 0.4615 0.5123 0.5123 0.5631 0.6139 0.5885 0.5377 0.4869 0.5885 0.6139 … ] (1×5000 double)} {[ 0.2383 0.3253 0.4124 0.4124 0.3834 0.3543 0.4414 0.4704 0.3834 0.4124 0.4704 0.5284 0.4994 0.4124 0.4124 0.4704 0.5284 0.4124 0.4124 0.4704 0.4704 0.3543 0.3253 0.3834 … ] (1×5000 double)} {[ -0.6125 -0.5854 -0.5854 -0.5311 -0.5311 -0.4768 -0.5854 -0.6397 -0.5311 -0.6125 -0.6397 -0.6397 -0.5583 -0.5583 -0.6397 -0.7211 -0.6397 -0.5854 -0.6668 -0.7754 -0.7211 … ] (1×5000 double)} {[ -0.3252 -0.2707 -0.2707 -0.3797 -0.3797 -0.3252 -0.2980 -0.3252 -0.3524 -0.2707 -0.2435 -0.2980 -0.4069 -0.3797 -0.2707 -0.3252 -0.3797 -0.3252 -0.2707 -0.2162 -0.2435 … ] (1×5000 double)}
signew
signew = 5×1 cell array
{[ -0.4456 -0.4074 -0.4120 -0.4502 -0.4883 -0.4876 -0.4868 -0.5441 -0.5708 -0.5968 -0.6349 -0.5677 -0.5280 -0.5303 -0.5876 -0.6220 -0.6258 -0.5494 -0.5693 -0.5998 -0.5616 … ] (1×7199 double)} {[-0.0973 -0.0444 -0.0014 0.0255 -0.0098 -0.0211 -0.0042 0.0664 0.1228 0.1630 0.1807 0.1821 0.1990 0.2357 0.3063 0.3345 0.3373 0.3549 0.3853 0.4206 0.4558 0.4911 0.5123 0.5123 … ] (1×7199 double)} {[ 0.2383 0.2987 0.3592 0.4124 0.4124 0.3987 0.3785 0.3584 0.4027 0.4486 0.4688 0.4148 0.3930 0.4140 0.4543 0.4946 0.5252 0.5051 0.4559 0.4124 0.4124 0.4462 0.4865 0.5268 … ] (1×7199 double)} {[ -0.6125 -0.5937 -0.5854 -0.5809 -0.5432 -0.5311 -0.5221 -0.4844 -0.5372 -0.5990 -0.6367 -0.5703 -0.5583 -0.6133 -0.6321 -0.6397 -0.6306 -0.5741 -0.5583 -0.5741 -0.6306 … ] (1×7199 double)} {[ -0.3252 -0.2874 -0.2707 -0.2798 -0.3555 -0.3797 -0.3706 -0.3328 -0.3101 -0.3048 -0.3237 -0.3426 -0.3252 -0.2700 -0.2510 -0.2662 -0.3101 -0.3858 -0.3933 -0.3585 -0.2828 … ] (1×7199 double)}
labl
labl = 5×1 cell array
{[n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a … ] (1×5000 categorical) } {[n/a P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P … ] (1×5000 categorical) } {[T T T T T T T T T T T T T T T T T T T T T T T T T T T T n/a n/a n/a … ] (1×5000 categorical) } {[n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a … ] (1×5000 categorical) } {[T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T … ] (1×5000 categorical)}
lablnew
lablnew = 5×1 cell array
{[n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a … ] (1×7199 categorical) } {[n/a P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P … ] (1×7199 categorical) } {[T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T … ] (1×7199 categorical)} {[n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a … ] (1×7199 categorical) } {[T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T … ] (1×7199 categorical)}
whos
Name Size Bytes Class Attributes ans 1x37 74 char c 4x1 432 cell cmdout 1x33 66 char fs 1x1 8 double fsnew 1x1 8 double i 1x1 8 double l 1x5000 5434 categorical labl 5x1 27690 cell lablnew 5x1 38685 cell lnew 1x7199 7633 categorical s 1x5000 40000 double sig 5x1 200520 cell signew 5x1 288480 cell snew 1x7199 57592 double t 1x5000 40000 double tnew 1x7199 57592 double
figure
subplot(211); plot(s)
subplot(212); plot(snew)
figure;
subplot(211); plot(double(l))
subplot(212); plot(double(lnew))
##### 5 件のコメント3 件の古いコメントを表示3 件の古いコメントを非表示
Chunru 2024 年 3 月 15 日
It is due to the way we produce new time:
t = (0:length(s)-1)/fs;
tnew = t(1):(1/fsnew):t(end);
We don't extend the time after the original end of the time. You can extend it if you like by add a small time after t(end).
Mibang 2024 年 3 月 15 日
Many thanks!

サインインしてコメントする。

### カテゴリ

Help Center および File ExchangeMultirate Signal Processing についてさらに検索

### Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by