[STATSGRASS] problem with RdbiPgSQL c.d.

[Roger Bivand]

On Wed, 30 Aug 2006, Jarek Jasiewicz wrote:

> Roger Bivand napisa?(a):
> > On Sun, 27 Aug 2006, Jarek Jasiewicz wrote:
> >
> >   
> >> on R, soon after I instaled sgeostat Rdbi cannot be loaded with message:
> >>     
> >
> > Not sgeostat, but probably version conflict in Rdbi. Please include the 
> > output of sessionInfo() after loading sgeostat, and any information you 
> > can see from the DESCRIPTION file in the directory with the installed Rdbi 
> > about builds and versions.
> >
> >   
> >>> library(Rdbi) 
> >>>       
> >>  
> >> Loading required package: Rdbi
> >> Loading required package: Rdbi
> >> Error in loadNamespace(package, c(which.lib.loc, lib.loc), keep.source = keep.source) :
> >>         cyclic name space dependencies are not supported
> >> Error: package 'Rdbi' could not be loaded
> >> Error: .onLoad failed in 'loadNamespace' for 'Rdbi'
> >> Error: package 'Rdbi' could not be loaded
> >>
> >>
> >> erlier Rdbi worked fine
> >>
> >> regards 
> >> jarek
> >>
> >>
> >> _______________________________________________
> >> statsgrass mailing list
> >> statsgrass at grass.itc.it
> >> http://grass.itc.it/mailman/listinfo/statsgrass
> >>
> >>     
> >
> >   
> Unfortunatly the problem is NOT solved..
> 
> The problem concering only RdbiPgSQL. sgeostat has no inlfluence... I 
> did the following:
> 
> 1) install R (2.3.1 from source)
> 2) install spgrass6, from your instruciton
> 3) install RdbiPgSQL according Marcus instruction
> 
> all successed.
> 
> 4) I load RdbiPgSQL library(RdbiPgSQL)... All OK
> 
> 5) do something... (lada data, manipulate and so on...
> 6 leave R without saving workspace
> 7) start R again load RdbiPgSQL ...All OK
> 8)do something (some objects are created) and leave R WITH SAVING workspace

Yes, this is the smoking gun. Please consider asking the Rdbi maintainer
who seems to be jzhang at jimmy.harvard.edu about this. 

I would first try to use the versions from
http://www.bioconductor.org/packages/1.8/bioc/ - I don't know why Markus
has 1.9, they are for R 2.4, not 2.3, because there is a major
discontinuity in object defintions between 2.3 and 2.4, which may be being
picked up when a package prepared for 2.4 tries to read an object saved
under 2.3.

Markus - was there a reason for writing 1.9? The change in object 
structure is very recent, and another one happened today - 2.4 is changing 
fast at the moment.

Roger

> 
> 9) load R again - ls() ...previous objects exit
> 10) try to load library(RdbiPgSQL): error
> 
> 
> Loading required package: Rdbi
> Loading required package: Rdbi
> Error in loadNamespace(package, c(which.lib.loc, lib.loc), keep.source = keep.source) :
>         cyclic name space dependencies are not supported
> Error: package 'Rdbi' could not be loaded
> Error: .onLoad failed in 'loadNamespace' for 'Rdbi'
> Error: package 'Rdbi' could not be loaded
> 
> ...and of course no function Rdbi works
> 
> 11)Again remove .Rdata from $home directory
> 12)start R again and library(RdbiPgSQL) works nice
> 
> so i thing some problem is with R workspace and information stored within.
> 
> Here is th "working" session info
> 
> Version 2.3.1 (2006-06-01)
> i686-pc-linux-gnu
> 
> attached base packages:
> [1] "methods" "stats" "graphics" "grDevices" "utils" "datasets"
> [7] "tcltk" "base"
> 
> other attached packages:
> spgrass6 rgdal maptools foreign gstat sp RdbiPgSQL Rdbi
> "0.3-1" "0.4-8" "0.5-12" "0.8-15" "0.9-31" "0.8-18" "1.7.2" "1.7.0"
>  >
> Regards
> Jarek Jasiewicz
> 

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no