Std deviation from fit (or other fitting tool)

This might be a previously answered question, but so far, I haven't found anything that specifically would help me.

How can I get the standard deviation for a fit using the fit function, or maybe some other fitting tool?

Thanks, R

Answer 1

[dpb]

I guess first is what you mean by "standard deviation for a fit"?

That said, however, from

>> help fit
 ...
[FO, G] = fit(X, Y, ...) returns appropriate goodness-of-fit measures, for
  the given inputs, in the structure G. G includes the fields:
      -- SSE         sum of squares due to error
      -- R2          coefficient of determination or R^2
      -- adjustedR2  degree of freedom adjusted R^2
      -- stdError    fit standard error or root mean square error
 ...
>>

Any of that of interest, perhaps???

[Ritu]

Hi,

Thanks for your answer. However, what I do need is this: I need to put the fitted values into another weighted fit using the cfit tool. And the weights of each point is defined as the inverse of the variance of that point. And to do this, I need the standard deviation of the point from the initial fit :| and that seems to be damned hard to get out of anything :/

Thanks though. :)

[dpb]

" I need the standard deviation of the point from the initial fit"

What, precisely, does this mean? A point has no standard deviation; it's easy enough to compute a residual at a point, but a variance? "Not so much..." which possibly explains why it's "damned hard to get out of anything". :)