Fitting data with uncertainties

I have experimntal measurements data points that have some uncertinties assosiated with them and I want to fit these data points to a curve where the values of the uncertinties are taken into considration.

My goal is to have an error value for when I substitute in the function generated from the fit. I want this error to take in consideration the fitting error and the uncertinity from the original data points as well in order to do a proper error propegation.

I need help on this regard for when I fit the data to gaussian desterbuation and totally different data I am fitting to a straight line.

for the gaussian disterbution I'm currently using:

    fit(x,y,'gauss1')

and for the staight line I'm using:

    robustfit(x,y)

BTW, I am not that good in stattistcs, so please go easy on me.