Central difference approximation scripts to calculate first derivatives of smoothed signal got from smooth function, Method: 'Savitzky-Golay'

I want to calculate first derivatives with respect to time for my smoothed data. I used Savitzky-Golay method in 'smooth' matlab function.I need information where I can get:

1. Any downloadable matlab scripts that can do this first derivative calculation by employing central difference approximation.

2. Any matlab function (in-built or user-written that can be downloaded) that can do the same exclusively for this smoothed result using Sgolay method in smooth function. Something like how 'fnder' is to 'csaps'.

Thanks in advance Harish

Answer 1

[Amit]

For central difference: http://www.mathworks.com/help/matlab/ref/gradient.html

For SG: http://www.mathworks.com/matlabcentral/fileexchange/30299-savitzky-golay-smoothdifferentiation-filters-and-filter-application

[SreeHarish Muppirisetty]

Hi

Thanks for the pointers.

Looks like gradient employs central difference approximation for calculating first derivatives.

How ever, i have a doubt on point 2. regarding the Sgolay method implementation.

Looks like the differentiation mentioned in the file is for fitting noisy data by variable polynomials, may be not for getting derivatives of the smoothed data. Am I correct?

Thanks again for your time

[Amit]

SG is a filter for noisy data. SG fits a pieces of data using polynomials order defined and smooths it. Thus, the derivative of the smoothed data will be in a way differential of those piecewise polynomials. In a way, you can get smoothed data and the derivative at the same time.

[SreeHarish Muppirisetty]

Great, Thanks!

Your answers are crisp and to the point!

Cheers!