For given column vector X and some values Y, FINDCLOSESTID2VAL returns the indices in X which are closest to Y, e.g.: FINDCLOSESTID2VAL([0.1 -1 -0.05 10],0) will return 3, since -0.05 is the closest value to 0.
This function is very handy for time series, where you want to get the id of a time point which might not be exactly matched.
This function is very fast for small to moderate sized 'X' and small 'Vals'. For a function doing the same work quickly on two very large arrays and with some additional features see NEARESTPOINT by Jos:
The figure on the upper left shows the max. length of 'Vals' versus the length of 'X' for which FINDCLOSESTID2VAL (f1) will be faster than NEARESTPOINT (f2) including the equation to compute the critical value for arbitrary lengths of 'X'.
David J. Mack (2021). Find the id of the closest value in an array to a given value (https://www.mathworks.com/matlabcentral/fileexchange/59043-find-the-id-of-the-closest-value-in-an-array-to-a-given-value), MATLAB Central File Exchange. Retrieved .
...but not completely: See speed comparison.
I accept the defeat ;-)
Nice wrapper around bsxfun. However, this makes it very slow for larger arrays, and increasingly so.
Alternatives, like my own function nearestpoint do the same job but much faster.
Inspired by: nearestpoint(x, y, m)
Find the treasures in MATLAB Central and discover how the community can help you!Start Hunting!