[R] Ways to work with R and Postgres

[Gabor Grothendieck]

2010/6/27 顾小波 <guxiaobo1982 at gmail.com>:
> Hi,
>
> I post this message to the general r-help list hoping anyone within a wider range have suggestions:
>
>
>
> There are three ways to integration R and postgres, especially on 64bit Microsoft windows Platform,
>
>
>
> 1. via RODBC package, which has 32 bit and 64 bit version for windows
>
> 2. via RPostgres interface, which only has 32bit version currently
>
> 3. via plr for Greenplum, which only supports a few kinds of functionality, and supports only specific versions of R.
>
>
>
> Do you have any idea about the advantages and disadvantages of each, and the differences among them
>

There is also the RpgSQL package.  In addition the sqldf package uses
RpgSQL.  sqldf by default uses SQLite but if the RpgSQL package is
loaded then it defaults to PostgreSQL.  Here BOD Is a built in R
data.frame:

> library(sqldf)
Loading required package: DBI
Loading required package: RSQLite
Loading required package: RSQLite.extfuns
Loading required package: gsubfn
Loading required package: proto
Loading required package: chron
> library(RpgSQL)
Loading required package: RJDBC
> BOD
  Time demand
1    1    8.3
2    2   10.3
3    3   19.0
4    4   16.0
5    5   15.6
6    7   19.8
> sqldf('select regr_slope(demand, "Time") slope,
+ regr_intercept(demand, "Time") intercept,
+ corr(demand, "Time") corr from BOD')
Loading required package: tcltk
Loading Tcl/Tk interface ... done
     slope intercept      corr
1 1.721429  8.521429 0.8030693

> coef(lm(demand ~ Time, BOD)); cor(BOD$Time, BOD$demand)
(Intercept)        Time
   8.521429    1.721429
[1] 0.8030693