In the operation of filtfilt, does the filtering process occur in the row direction or the column direction?

I have a question regarding the direction in which filtfilt operates. Time series data is stored for each column, such as in data.mat. I am considering filtering this data separately for each column. My question has two parts:

• In the operation of filtfilt, does the filtering process occur in the row direction or the column direction?
• If it operates in the row direction, is there a way to make it operate in the column direction?

I appreciate your response.

load data.mat
sampling_rate = 200;
cutoff_frequency = 12;
filter_order = 4;
[b, a] = butter(filter_order, cutoff_frequency / (0.5.*sampling_rate), 'low');
filt_data = [data(:,1:2) filtfilt(b, a, data(:,3:end))]

filt_data = 1049×155

1.0e+03 * 0.0010 0 -0.3698 1.2958 0.1429 -0.3810 1.2850 -0.1730 -0.3066 0.8857 0.1010 -0.2512 0.8425 0.0766 -0.3297 0.7808 0.1221 -0.2978 0.8054 0.1134 -0.3281 0.6877 0.1396 -0.3339 0.5977 0.1380 -0.3439 0.5118 0.1294 -0.2394 0.0020 0.0000 -0.3700 1.2958 0.1429 -0.3811 1.2850 -0.1731 -0.3066 0.8856 0.1010 -0.2512 0.8425 0.0766 -0.3297 0.7808 0.1221 -0.2978 0.8054 0.1134 -0.3280 0.6877 0.1397 -0.3339 0.5977 0.1380 -0.3438 0.5118 0.1295 -0.2394 0.0030 0.0000 -0.3701 1.2959 0.1429 -0.3812 1.2851 -0.1731 -0.3067 0.8856 0.1009 -0.2511 0.8425 0.0766 -0.3297 0.7808 0.1221 -0.2978 0.8054 0.1135 -0.3278 0.6878 0.1397 -0.3339 0.5977 0.1380 -0.3437 0.5118 0.1295 -0.2394 0.0040 0.0000 -0.3702 1.2959 0.1429 -0.3814 1.2851 -0.1732 -0.3067 0.8856 0.1008 -0.2511 0.8425 0.0766 -0.3297 0.7808 0.1221 -0.2978 0.8054 0.1135 -0.3277 0.6878 0.1397 -0.3339 0.5977 0.1380 -0.3436 0.5117 0.1295 -0.2393 0.0050 0.0000 -0.3703 1.2959 0.1429 -0.3815 1.2851 -0.1732 -0.3068 0.8856 0.1008 -0.2511 0.8425 0.0766 -0.3296 0.7808 0.1221 -0.2977 0.8054 0.1135 -0.3276 0.6878 0.1397 -0.3338 0.5978 0.1380 -0.3436 0.5117 0.1296 -0.2393 0.0060 0.0000 -0.3704 1.2959 0.1429 -0.3816 1.2851 -0.1733 -0.3069 0.8856 0.1008 -0.2511 0.8425 0.0766 -0.3296 0.7808 0.1221 -0.2977 0.8054 0.1135 -0.3275 0.6879 0.1398 -0.3338 0.5978 0.1380 -0.3435 0.5117 0.1297 -0.2393 0.0070 0.0000 -0.3705 1.2959 0.1428 -0.3817 1.2852 -0.1734 -0.3069 0.8856 0.1007 -0.2511 0.8425 0.0766 -0.3296 0.7808 0.1221 -0.2977 0.8054 0.1135 -0.3274 0.6879 0.1398 -0.3338 0.5978 0.1380 -0.3434 0.5117 0.1297 -0.2393 0.0080 0.0000 -0.3706 1.2959 0.1428 -0.3818 1.2852 -0.1734 -0.3070 0.8856 0.1007 -0.2511 0.8425 0.0766 -0.3296 0.7808 0.1221 -0.2977 0.8054 0.1135 -0.3273 0.6879 0.1398 -0.3337 0.5978 0.1380 -0.3434 0.5117 0.1298 -0.2392 0.0090 0.0000 -0.3706 1.2959 0.1427 -0.3819 1.2853 -0.1735 -0.3071 0.8856 0.1007 -0.2511 0.8425 0.0766 -0.3296 0.7808 0.1221 -0.2977 0.8054 0.1135 -0.3273 0.6879 0.1398 -0.3337 0.5978 0.1380 -0.3433 0.5117 0.1298 -0.2392 0.0100 0.0000 -0.3707 1.2959 0.1427 -0.3820 1.2853 -0.1735 -0.3072 0.8856 0.1007 -0.2512 0.8425 0.0765 -0.3296 0.7808 0.1221 -0.2978 0.8054 0.1135 -0.3272 0.6880 0.1398 -0.3336 0.5978 0.1380 -0.3433 0.5117 0.1298 -0.2392

Answer 1

[the cyclist]

Per the documentation page filtfilt ...

"The function operates along the first array dimension of x unless x is a row vector. If x is a row vector, then the function operates along the second dimension."

"First array dimension" means down the columns.

[Walter Roberson]

And there is no option to filter along rows when the data is not a row vector. If you need to do that, then transpose, .' or permute before filtfilt and convert back afterwards.

[Sho]

Thank you so much for your help on my MATLAB question. Your response helped me solve my problem.