How to model multiple categorical variable in r, using gee model

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

How to model multiple categorical variable in r, using gee model

augustinus ntjamba
Good morning.
I'm a student at present working on my final year project.
Kindly asking for help on how to model longitudinal categorical data.

In my data set I have the following variables :type of crime,year,   month,
date and time.treating type of crime as the response variable and there's
12 levels  (Type of crime), while the rest of the variables are
independent.
What model will best fit my data?

I have tried using geeglm And this does show differences in correlation
matrix that should be selected as the best model, secondly I tried using
multgee package "multLORgee" which never have me outputs and lastly I tried
using multnom the function returns the same AIC in the working correction
matrix,
How do i solve this problem
Thank you.

        [[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: How to model multiple categorical variable in r, using gee model

Jim Lemon-4
Hi Augustinius,
You have been set a problem that requires a lot more information than
is in your request. Also, you have flagged it as a "homework" problem,
so you are unlikely to get much help on this list. Sadly, this sort of
problem sometimes arises when being nice to the instructor is more
important than having the relevant knowledge and I can't help you with
that.

Jim

On Mon, Sep 21, 2020 at 10:39 PM augustinus ntjamba
<[hidden email]> wrote:

>
> Good morning.
> I'm a student at present working on my final year project.
> Kindly asking for help on how to model longitudinal categorical data.
>
> In my data set I have the following variables :type of crime,year,   month,
> date and time.treating type of crime as the response variable and there's
> 12 levels  (Type of crime), while the rest of the variables are
> independent.
> What model will best fit my data?
>
> I have tried using geeglm And this does show differences in correlation
> matrix that should be selected as the best model, secondly I tried using
> multgee package "multLORgee" which never have me outputs and lastly I tried
> using multnom the function returns the same AIC in the working correction
> matrix,
> How do i solve this problem
> Thank you.
>
>         [[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: How to model multiple categorical variable in r, using gee model

Jim Lemon-4
Hi Augustinius,
You are probably familiar with some of these:

http://finzi.psych.upenn.edu/R/library/geepack/html/geeglm.html
https://faculty.washington.edu/heagerty/Courses/b571/homework/geepack-paper.pdf
https://rdrr.io/cran/geex/f/vignettes/articles/mestimation_bib.Rmd

Good luck with it.

Jim

On Tue, Sep 22, 2020 at 9:52 AM augustinus ntjamba
<[hidden email]> wrote:

>
> even a source perhaps that you may refer me too, so I get some idea's tht will be appreciated as well.
>
> On Tue, 22 Sep 2020, 01:50 augustinus ntjamba <[hidden email]> wrote:
>>
>> Thank you for the feedback.
>> actually its not a "home work", it's a "project" that I'm working on and I'm challenged as which function in generalized estimating equations to use, i try using geeglm ,multgee, multinom to model the data, however; I'm unable to determine as which working correction matrix will better fit the model (model selection ) for the fact that all the model with different correlation structure keeps returning the same outputs (I do not observe an difference) hence my concern as it supposed not be the case.
>>
>> This is one of the reason I'm asking for help, like how do I go about it?
>> Your gguidelines will be highly appreciated.
>> My regards,

______________________________________________
[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: How to model multiple categorical variable in r, using gee model

augustinus ntjamba
Thanks for coming through sir. I will check it out.

On Tue, Sep 22, 2020 at 2:47 AM Jim Lemon <[hidden email]> wrote:

> Hi Augustinius,
> You are probably familiar with some of these:
>
> http://finzi.psych.upenn.edu/R/library/geepack/html/geeglm.html
>
> https://faculty.washington.edu/heagerty/Courses/b571/homework/geepack-paper.pdf
> https://rdrr.io/cran/geex/f/vignettes/articles/mestimation_bib.Rmd
>
> Good luck with it.
>
> Jim
>
> On Tue, Sep 22, 2020 at 9:52 AM augustinus ntjamba
> <[hidden email]> wrote:
> >
> > even a source perhaps that you may refer me too, so I get some idea's
> tht will be appreciated as well.
> >
> > On Tue, 22 Sep 2020, 01:50 augustinus ntjamba <
> [hidden email]> wrote:
> >>
> >> Thank you for the feedback.
> >> actually its not a "home work", it's a "project" that I'm working on
> and I'm challenged as which function in generalized estimating equations to
> use, i try using geeglm ,multgee, multinom to model the data, however; I'm
> unable to determine as which working correction matrix will better fit the
> model (model selection ) for the fact that all the model with different
> correlation structure keeps returning the same outputs (I do not observe an
> difference) hence my concern as it supposed not be the case.
> >>
> >> This is one of the reason I'm asking for help, like how do I go about
> it?
> >> Your gguidelines will be highly appreciated.
> >> My regards,
>

        [[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.