X is rank deficient to within machine precision.

[Michele Costola]

Hi everybody,
I'm doing regression on my y with x serie with 15 lag, so I have 17 column (const+1+15) in my x. When I start the regression I get this error:

Warning: X is rank deficient to within machine precision.
> In regress at 82

I tried to search on the newsreader but I don't really understand what could be the problem and how to solve it.
Thank you for your support,
Michele

[Peter Perkins]

Michele Costola wrote:
> Hi everybody,
> I'm doing regression on my y with x serie with 15 lag, so I have 17 column (const+1+15) in my x. When I start the regression I get this error:
>
> Warning: X is rank deficient to within machine precision.
>> In regress at 82

Most likely you have too many predictor variables, and too few observations. As an alaogy, imagine trying to fit a cubic polynomial regression with only two observations. REGRESS will make a choise about which coefficients to set to zero, but it can't possibly know what you really want.

Hope this helps.

[FG]

This is likely multicollinearity. if you the variables are dummies, remember to leave out one of the dummies as the model's intercept will pick up the dummy you leave out as the base model.


Peter Perkins <Peter....@MathRemoveThisWorks.com> wrote in message <hcmmlt$lqh$2...@fred.mathworks.com>...

[Dinakar]

You have an extra intercept - you dont need both const and 1. Drop one and see.

Regards
Dj

"mklcst mklcst" wrote in message <hch1ji$1g0$1...@fred.mathworks.com>...