Thread Subject: fitted values

Command for printing out the fitted values of a basic OLS regression?

"Janis Weir" <weij@bankofcanada.ca> wrote in message <hf0tbg$dju$1@fred.mathworks.com>...
> Command for printing out the fitted values of a basic OLS regression?

You have the statistics toolbox, see

>>doc regress

Hope that helps,
Wayne

"Wayne King" <wmkingty@gmail.com> wrote in message <hf0u0g$p9b$1@fred.mathworks.com>...
> "Janis Weir" <weij@bankofcanada.ca> wrote in message <hf0tbg$dju$1@fred.mathworks.com>...
> > Command for printing out the fitted values of a basic OLS regression?
>
> You have the statistics toolbox, see
>
> >>doc regress
>
> Hope that helps,
> Wayne

I see the regress command, but not the argument to have matlab display the fitted values?

Hi Janis



If you look at the documentation section for regress in Statistics Toolbox,
you'll find the following example code



% Load data on cars; identify weight and horsepower as predictors, mileage
as the response:



load carsmall

x1 = Weight;

x2 = Horsepower; % Contains NaN data

y = MPG;



% Compute regression coefficients for a linear model with an interaction
term:



X = [ones(size(x1)) x1 x2 x1.*x2];

b = regress(y,X) % Removes NaN data

b =

  60.7104

  -0.0102

  -0.1882

   0.0000





The default output from the regress command is a matrix of coefficients.



Regards,



Richard




"Janis Weir" <weij@bankofcanada.ca> wrote in message
news:hf0tbg$dju$1@fred.mathworks.com...
> Command for printing out the fitted values of a basic OLS regression?

"Richard Willey" <rwilley@mathworks.com> wrote in message <hf12nn$k2m$1@fred.mathworks.com>...
> Hi Janis
>
>
>
> If you look at the documentation section for regress in Statistics Toolbox,
> you'll find the following example code
>
>
>
> % Load data on cars; identify weight and horsepower as predictors, mileage
> as the response:
>
>
>
> load carsmall
>
> x1 = Weight;
>
> x2 = Horsepower; % Contains NaN data
>
> y = MPG;
>
>
>
> % Compute regression coefficients for a linear model with an interaction
> term:
>
>
>
> X = [ones(size(x1)) x1 x2 x1.*x2];
>
> b = regress(y,X) % Removes NaN data
>
> b =
>
> 60.7104
>
> -0.0102
>
> -0.1882
>
> 0.0000
>
>
>
>
>
> The default output from the regress command is a matrix of coefficients.
>
>
>
> Regards,
>
>
>
> Richard
>
>
>
>
> "Janis Weir" <weij@bankofcanada.ca> wrote in message
> news:hf0tbg$dju$1@fred.mathworks.com...
> > Command for printing out the fitted values of a basic OLS regression?
> I'm using the result=ols(y,x) command. I assume that the result.yhat command prints the same estimate coefficients?

>> "Janis Weir" <weij@bankofcanada.ca> wrote in message
>> news:hf0tbg$dju$1@fred.mathworks.com...
>> > Command for printing out the fitted values of a basic OLS regression?
>> I'm using the result=ols(y,x) command. I assume that the result.yhat
>> command prints the same estimate coefficients?


Hi Janis


We (The MathWorks) don't have a command named OLS.



This is only a guess, however, I'm assuming that one of your colleagues at
the Bank of Canada has designed their own custom regression function.
(Alternatively, I know that there is a file exchange submission named OLS,
however, this is performing Orthogonal Least-Squares rather than Ordinary
Least-Squares)



You should have a file names "ols.m" in one of your directories. If you
open this up in the MATLAB editor you might be able to find out whats going
on (with luck, the individual who wrote the function might also have
provided some documentation and/or examples)



I'm not sure what result.yhat will return. This depends completely on how
the function was written.



regards,



Richard