I've just spent two hours finding out, what is the meaning of the sig array in mrqmin function. :mad:

Since I have only two one-dimensional data arrays (x and y) an explanation: "individual standard deviations sig[1..ndata]" is completely unclear for me (hope that for the authors it is :p ).


Does anybody know what the sig is??

The sig[1..ndata] array is your estimation of the standard deviation of the input data. Specifically, I think it should be your estimation of the standard deviation of the y data. In the simplest case you might choose a value that you think is reasonable and make it the same for each y value. A more sophisticated approach is to repeat your experiment n times so that you have n y values for each x data point and then estimate the standard deviation for each set of n y values. In this approach you would probably want to use the mean values of y (for each x data point) as your y data.

Hi did u implement the mrqmin? if for example I have a y array with the following values [1,2,3,4,5,6,7]. Shouldn't the standard deviation of each one of them be.

mean= 28/7=4
SD[0]=(1-3)^2= 4
SD[1]=(2-3)^2=1
SD[2]=(3-3)^2=0
and so on......

Am I wrong here?