Début du message réexpédié :
> De: Depiereux Constant <constant.depiereux@aqte.be>
> Date: 16 mars 2005 00:11:11 GMT+01:00
> À: spjgmwn@iop.kcl.ac.uk
> Objet: Rép : [R] RODBC, sqlSave and sqlAppend
>
>
> Le 15-mars-05, à 16:19, Matthew W Nash a écrit :
>
>> Hi all,
>>
>> I am currently trying to read, write and append data between R and MS
>> access
>> using the RODBC library functions. I have no problems reading in the
>> data
>> but when using sqlSave and sqlAppend it doesn't seem to work. I
have
>> made
>> sure that all the column names are sensible and there are no gaps etc
>> in the
>> variables. My call looks like this:
>>
>> sqlSave(channel,treatlist,test=F)
>>
>> I've played with various options but what consistently happens is
>> that R
>> writes a new table (column names a written and sensible variable
>> types are
>> assigned), but it doesn't actually write any data into it. When I
run
>> sqlAppend there are no error messages, but when I look at the MS
>> access
>> database nothing has been written. (I always have the database closed
>> when
>> doing this.)
>>
>> What am I doing wrong?
>>
>> Matthew Nash,
>>
>> Post-Doctoral Research Worker,
>> GENDEP study,
>> SGDP, Institute of Psychiatry,
>> PO82, Room CB.15,
>> 16 De Crespigny Park,
>> London, SE5 8AF,
>> United Kingdom
>>
>> Phone: (+44) 207 848 0805
>>
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> ~~~~~~
>> ~
>> Announcing the 6th MRC SGDP SUMMER SCHOOL 25 - 29 July 2005 with
>> courses in
>> a) Twin model fitting, Mx
>> b) Microarrays (Affymetrix), gene expression, SNPs
>> c) Linkage, association and allied methods
>> http://sgdp.iop.kcl.ac.uk/summerschool/
>>
>> ______________________________________________
>> R-help@stat.math.ethz.ch mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide!
>> http://www.R-project.org/posting-guide.html
>>
>>
>
>
> hello,
>
>
> I have also been hardly struggling with the same problem, experiencing
> repetitive R and windows crashes.
>
> I finally succeeded to make it work and discovered that although RODBC
> works extremely well with , most of the problems are due to the ODBC
> mechanism only.
>
> Please find herewith some of the stupid things that I have learned in
> the hard way.
>
> 1. In creating a dataframe, avoid variable such as
'variablename[n]'
> or 'dataframename$component'. Using them collapse everything.
>
> 2. Dont create table description from SQL and pgAdmin or the like,
> let R/RODBC create it thru sqlSave, using first 'append=FALSE' and
> 'safer=FALSE' options
>
> 3. In pgAdmin, adjust cautiously the data type, even if you have to
> accept missdimentionned integer or float (the price of disk is
> unsignificant today)
>
>
> I realize its look like vaudoo and black magic, but shoulf I have done
> that from the start, I would have save a long day of work debugging
> phantoms, ...
>
> If some body has better practices, there would be warmly wellcomed.
>
> Sorry for disturbance and naive approach.
>
> Best regards
>
>
> Constant Depièreux
> Managing Director
> Applied Quality Technologies Europe sprl
> Rue des Déportés 123, B4800 Verviers
> Tel : + 32 87 29 21 75
> Fax : +32 87 29 21 71
> Mobile : +32 475 555 818
> e-Mail : constant.depiereux@aqte.be
> Web presence : http://www.aqte.be
> Skype : cdepiereux
>
>
Constant Depièreux
Managing Director
Applied Quality Technologies Europe sprl
Rue des Déportés 123, B4800 Verviers
Tel : + 32 87 29 21 75
Fax : +32 87 29 21 71
Mobile : +32 475 555 818
e-Mail : constant.depiereux@aqte.be
Web presence : http://www.aqte.be
Skype : cdepiereux
[[alternative text/enriched version deleted]]
