Statistical tests on residuals

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Statistical tests on residuals

NehaBologna
Hi

I have continuous data I.e regression based. If I have residuals for
algorithm 1 and algorithm 2 like this,

X=resid(alg1) and y=resid (alg2)

Can we perform the Wilcoxon test as.

Wilcox.test(x,y)

Thanks

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Statistical tests on residuals

Rui Barradas
Hello,

Inline.

Às 16:28 de 29/12/19, Neha gupta escreveu:

> Hi
>
> I have continuous data I.e regression based. If I have residuals for
> algorithm 1 and algorithm 2 like this,
>
> X=resid(alg1) and y=resid (alg2)
>
> Can we perform the Wilcoxon test as.
>
> Wilcox.test(x,y)

Probably not, R is case sensitive and without loading an external
package that instruction will give an error. Try

wilcox.test(x, y)


with *lower* case *w*.

By the way, why would you test the residuals of regression fits?
It doesn't make sense, especially if the models (algorithms) assume
residuals with mean zero. I suggest you learn more about the techniques
you are using to model your data before trying to test.


Hope this helps,

Rui Barradas


>
> Thanks
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> [hidden email] mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Statistical tests on residuals

Rui Barradas
Hello,

Please cc the list.

Yes, that's what I mean.
Besides, the code as provided doesn't produce an object (search) with a
resid method. Are you sure resid(search) runs without errors?

Hope this helps,

Rui Barradas

Às 16:55 de 29/12/19, Neha gupta escreveu:

> Hello Rui,
>
> I use caret and CART
>
> search <- train(Results ~ ., data = training,
>                       method = "rpart",
>
>                       metric = "MAE",
>                       preProc = c("center", "scale", "zv" ),
>
> getTrainPerf(search)
> x=resid(search)
>
> You mean its not the right way for statistical test?
>
> On Sun, Dec 29, 2019 at 5:41 PM Rui Barradas <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Hello,
>
>     Inline.
>
>     Às 16:28 de 29/12/19, Neha gupta escreveu:
>      > Hi
>      >
>      > I have continuous data I.e regression based. If I have residuals for
>      > algorithm 1 and algorithm 2 like this,
>      >
>      > X=resid(alg1) and y=resid (alg2)
>      >
>      > Can we perform the Wilcoxon test as.
>      >
>      > Wilcox.test(x,y)
>
>     Probably not, R is case sensitive and without loading an external
>     package that instruction will give an error. Try
>
>     wilcox.test(x, y)
>
>
>     with *lower* case *w*.
>
>     By the way, why would you test the residuals of regression fits?
>     It doesn't make sense, especially if the models (algorithms) assume
>     residuals with mean zero. I suggest you learn more about the techniques
>     you are using to model your data before trying to test.
>
>
>     Hope this helps,
>
>     Rui Barradas
>
>
>      >
>      > Thanks
>      >
>      >       [[alternative HTML version deleted]]
>      >
>      > ______________________________________________
>      > [hidden email] <mailto:[hidden email]> mailing list
>     -- To UNSUBSCRIBE and more, see
>      > https://stat.ethz.ch/mailman/listinfo/r-help
>      > PLEASE do read the posting guide
>     http://www.R-project.org/posting-guide.html
>      > and provide commented, minimal, self-contained, reproducible code.
>      >
>

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Statistical tests on residuals

NehaBologna
Hello Rui, thanks for your input

Yes, it runs. In fact, it gives me the result like below: (I just copied
few, the list is high)

   X2           X3           X4           X5           X6           X7
      X8
  -95.835556  -182.235556  -177.435556  -188.235556  -205.035556
 -202.635556   139.364444
         X10          X11          X16          X17          X21
 X22          X23
 -141.435556  -189.435556   146.564444   110.564444  -153.435556
 86.564444   -93.435556
         X24          X25          X27          X28          X30
 X31          X32
 -123.435556    -3.435556  -143.435556    25.564444  -151.435556
-43.435556   -2

On Sun, Dec 29, 2019 at 6:20 PM Rui Barradas <[hidden email]> wrote:

> Hello,
>
> Please cc the list.
>
> Yes, that's what I mean.
> Besides, the code as provided doesn't produce an object (search) with a
> resid method. Are you sure resid(search) runs without errors?
>
> Hope this helps,
>
> Rui Barradas
>
> Às 16:55 de 29/12/19, Neha gupta escreveu:
> > Hello Rui,
> >
> > I use caret and CART
> >
> > search <- train(Results ~ ., data = training,
> >                       method = "rpart",
> >
> >                       metric = "MAE",
> >                       preProc = c("center", "scale", "zv" ),
> >
> > getTrainPerf(search)
> > x=resid(search)
> >
> > You mean its not the right way for statistical test?
> >
> > On Sun, Dec 29, 2019 at 5:41 PM Rui Barradas <[hidden email]
> > <mailto:[hidden email]>> wrote:
> >
> >     Hello,
> >
> >     Inline.
> >
> >     Às 16:28 de 29/12/19, Neha gupta escreveu:
> >      > Hi
> >      >
> >      > I have continuous data I.e regression based. If I have residuals
> for
> >      > algorithm 1 and algorithm 2 like this,
> >      >
> >      > X=resid(alg1) and y=resid (alg2)
> >      >
> >      > Can we perform the Wilcoxon test as.
> >      >
> >      > Wilcox.test(x,y)
> >
> >     Probably not, R is case sensitive and without loading an external
> >     package that instruction will give an error. Try
> >
> >     wilcox.test(x, y)
> >
> >
> >     with *lower* case *w*.
> >
> >     By the way, why would you test the residuals of regression fits?
> >     It doesn't make sense, especially if the models (algorithms) assume
> >     residuals with mean zero. I suggest you learn more about the
> techniques
> >     you are using to model your data before trying to test.
> >
> >
> >     Hope this helps,
> >
> >     Rui Barradas
> >
> >
> >      >
> >      > Thanks
> >      >
> >      >       [[alternative HTML version deleted]]
> >      >
> >      > ______________________________________________
> >      > [hidden email] <mailto:[hidden email]> mailing list
> >     -- To UNSUBSCRIBE and more, see
> >      > https://stat.ethz.ch/mailman/listinfo/r-help
> >      > PLEASE do read the posting guide
> >     http://www.R-project.org/posting-guide.html
> >      > and provide commented, minimal, self-contained, reproducible code.
> >      >
> >
>

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.