topic Re: Solve for coefficient in intergral function in New SAS User

Solve for coefficient in intergral function

tampham92 — Fri, 25 Jan 2019 20:24:53 GMT

Hi,

I have monthly data on X and Y since 2005 till now.

I would like to solve for the coefficient (b and a) of regressing y on x using the equation below. I came across a function called PROC NLMIXED in SAS but I am not quite sure that is a right function for me to use to solve for b and a in the following equation. Thanks a lot for your help!

X in the case is bounded between 0 and c

Re: Solve for coefficient in intergral function

PGStats — Fri, 25 Jan 2019 23:08:18 GMT

To fit a non-linear function Y = f(X) by regression to some X and Y data, you can use proc nlin. If the model linking Y to X involves random parameters, other than pure random error, you should use proc nlmixed.

Start with proc nlin. It is fairly simple to use. The only tricky part is finding some good starting values for the parameters. Read the documentation.

https://documentation.sas.com/?docsetId=statug&docsetVersion=14.3&docsetTarget=statug_nlin_overview.htm&locale=en

Don't hesitate to ask for more help if you encounter problems along the way.

Re: Solve for coefficient in intergral function

tampham92 — Mon, 28 Jan 2019 13:18:29 GMT

Would I be able to use the regular proc model to compute the coefficients to fit the data I have using defined equations.