I have a sample of Lokomat walking therapy data. the data consis of 75000x4 time sequence data for 5 minutes of recording (refer excel file); time - column 1, Gait index - column 2, hip angle - column 3, knee angle - column 4. The plots for whole recording are as in the pdf file; about 100 cycles. I plotted for a few cycles to see the patterns.

Gait cycle identification can already be viewed from the gait index that repeats from 0-1000 (ramp pattern) for each gait cycle. It does not need feature extraction to determine the cycle because the gait index can be used to determine a cycle.

How to segment each cycle (get the data for each cycle) then get average cycle while every cycle has different numbers of data? For each cycle the data counts is different, so to find the average, need to interpolate each cycle data to the same number of data (101 points - 0% -100%). How to use MATLAB to do the operations? How can I separate the data into 70% to find the average and another 30% to test it?