Constrained Dynamic Time Warping Distance Measure

This MEX function computes the constrained dynamic time warping distance between two time series.

This is the C source code for a Matlab executable (MEX) function named cdtw_dist(). This function returns the Constrained (by Sakoe-Chiba band) Dynamic Time Warping distance between two time series. This function has 3 input parameters and 1 return value: "distance = cdtw_dist(query_vector, candidate_vector, radius);". The distance value is the square root of the minimum sum of squared differences divided by the number of comparisons. If radius == Inf, distance is the unconstrained dynamic time warping distance. Both the time and space complexity of this function are O(mn), where m is the length of the query_vector and n is the length of the candidate_vector. This MEX function was tested using Matlab (R14SP1) for Windows with Matlab's Lcc C (v2.4). These Matlab commands were used to compile this function: "mex -setup; mex cdtw_dist.c;".