[R-sig-Epi] R-sig-Epi Digest, Vol 45, Issue 1
BXC (Bendix Carstensen)
bxc at steno.dk
Sat Sep 4 09:36:42 CEST 2010
Further to this thread:
It seems that the only shortcoming in R in relation to epidemiological practise is the data entry side. Which can be excellently handled by EpiData, which can output Stata data files (.dta) that are seamlessly imported to R via read.dta() from the foreign package.
I doubt there are anything on the analysis side in EpiInfo which is not in R, so the only issue is really the accessibility where the learning curve for R (or any command driven package) is steep.
But this is really just a feature of science: You must be able to reproduce your results, so you have to generate code which will re-do your analyses. So the point-and-click systems should not be used in scientific reporting anyway. Some of these systems leave a trace, but the code left is normally largely impenetrable, so not really a documentation of anything but the time you spent operating the mouse.
Best regards,
Bendix
Epi maintainer
> -----Original Message-----
> From: r-sig-epi-bounces at stat.math.ethz.ch
> [mailto:r-sig-epi-bounces at stat.math.ethz.ch] On Behalf Of
> Pedro Emmanuel Alvarenga Americano do Brasil
> Sent: 4. september 2010 05:47
> To: r-sig-epi at stat.math.ethz.ch
> Subject: Re: [R-sig-Epi] R-sig-Epi Digest, Vol 45, Issue 1
>
> Kevin,
>
> Although I thought at first it sound like a nice idea it also
> seems a tough task!!! Unfortunately Im a epidemiologist and I
> do not fell able to develop complex stuff in R.
>
> I did use very few times EpiInfo long ago and in the last
> four year I invested a lot of time in R due to my PhD project.
>
> I wondered about some points in this idea of yours.
> R is a very powerful and complex tool, so what is epiinfo
> able to do that R is not?
> In the other way around, what is R is able to do that epiInfo is not?
>
> My guess is that R is able to do most, if not all, analysis
> epiinfo can.
> Thus what are the weak points? EpiInfo is directod to users (most
> epidemiologist) which may have basics to moderate
> understanding of statistics and like to freindly data entry
> and dropdown menus and dialog boxes with more intuitive
> interface and simple and friendly documentation, and many
> options may be confusing, right?
>
> While R is command line software in nature. My own experience
> was like banging my head all over for more than a year until
> I felt myself familiar enough to not be scared and find ways
> on my own.
>
> Thus it very very short it would be like... make a package
> called EpiInfo that have the strong points of R and the
> strong points of EpiInfo. This would include a GUI for R.
> There are many GUIs for R around for multi and specific
> plataform such as Rcommander, Rpad, JGR, VIM, RkWard,
> SciViews etc http://www.sciviews.org/_rgui/
>
> So ... what would be the advantages for the end user (and also for
> developers) worth of such effort?
>
> (1) developing a package would essentially concentrating
> efforts in the analysis tool and very few, if any, effort in
> the the software core. Core team already made a software
> fully compatible with any OS. In very very short it would be
> an Rcommander with menus for the epi packages already
> avaliable (it couldnt be so simple) :)
>
> (2) make a friendly and yet powerfull tool for
> epidemiologists not initiated in R.
>
> (3) As far as know R does not have a data entry tool. Very
> few functions such as data.entry allow to type in data in a
> spreadsheet layout. Nothing similar EpiInfo, EPi6 or EpiData.
>
> In this particular matter (point 3) this package would very
> encouraging for the not initiated to use R.
>
> Of course, I can not speak for the core R team neither for
> the four epi packages developers already available, but if I
> was you I would consider make Rcommander plug-in or a GUI for
> the epi packages as a first step.
>
> Kind regards and good luck,
>
> Abraço forte e que a força esteja com você,
>
> Dr. Pedro Emmanuel A. A. do Brasil
> Instituto de Pesquisa Clínica Evandro Chagas Fundação Oswaldo
> Cruz Rio de Janeiro - Brasil Av. Brasil 4365 Tel 55 21 3865-9648
> email: pedro.brasil at ipec.fiocruz.br
> email: emmanuel.brasil at gmail.com
>
> ---Apoio aos softwares livres
> www.zotero.org - gerenciamento de referências bibliográficas.
> www.broffice.org ou www.openoffice.org - textos, planilhas ou
> apresentações.
> www.epidata.dk - entrada de dados.
> www.r-project.org - análise de dados.
> www.ubuntu.com - sistema operacional
>
> [[alternative HTML version deleted]]
>
>
More information about the R-sig-Epi
mailing list