Re: Fit with constraints?

*To*: mathgroup at smc.vnet.net*Subject*: [mg23353] Re: Fit with constraints?*From*: "Martin Rommel" <Martin_Rommel at mac.com>*Date*: Thu, 4 May 2000 02:59:20 -0400 (EDT)*Organization*: Semitest*References*: <8eg5ee$giq@smc.vnet.net> <8ekvpr$qb0@smc.vnet.net>*Sender*: owner-wri-mathgroup at wolfram.com

> http://support.wolfram.com/Math/Statistics/NonlinearOptimization.html > > may have the information you want. No, it doesn't. I checked the website (at least superficially). The answer which is as simple as I expected (still I did not think of it) came from Dave Withoff of Wolfram Research. >Fixing a point or setting an intercept just means using a model >that always passes through a particular point, as in > >In[1]:= Fit[{{1,2},{3,7},{4,6},{7,9},{11,14}}, x-4, x] > >Out[1]= 1.64706 (-4 + x) > >which fits the data to a straight line that passes through 4 (that is, >it sets the y intercept to 4). If you have an example where you're >not sure how to do this, you could try posting your example, or >contacting technical support. In any case, fixing a point is just >a simple modification of the model in the input to Fit. > >Dave Withoff >Wolfram Research