GreenplumR is a R client that designed for Greenplum Database (> 6.0). With GreenplumR installed in R environment, users can interact with data in Greenplum Database for analytics purpose. GreenplumR provides a rich interface to allow user access both tables and views with minimal data transfer. Users can easily access database without any knowledge of SQL. Moreover, GreenplumR can allow user to execute their own R code combine with data in Greenplum Database directly via built-in APPLY functions. GreenplumR can work with PL/Conatiner to provide a high preformance sandbox R runtime environment.
Pre-Installation:
You need to install ini, shiny, RPostgreSQL packages before install GreenplumR.
install.packages(c("ini", "shiny", "RPostgreSQL"))
You can install GreenplumR through 2 ways:
- Install throught devtools and get the latest development version from github by running the following code (need R >= 3.6.0):
## install.packages("devtools")
devtools::install_github("GreenplumR", "greenplum-db")
- Or You can download the source tarball directly from here, and then install the tarball in R:
install.packages("greenplum-db-GreenplumR-xxxx.tar.gz", repos = NULL, type = "source")
library(GreenplumR)
## connect to a local database
db.connect(host = "localhost",port = 5432, dbname = "test")
## connect to database via DSN
db.list() # list connections
db.objects() # list all tables in connection 1
db.objects("tbl.tbl") # search table in connection 1
#[1] "tbl.tbl"
db.objects("TBL.TBL") # NOTICE: unlike PG's behavior, db.objects in R is a Case-Sensitive function, the table name in greenplum if tbl.tbl, so this function call's result is NULL
#NULL
## wrapper of a table in connection 1
t <- db.data.frame("tbl.tbl")
t
## example of gpapply
fn.function_plus_one <- function(num)
{
return (num[[1]] + 1)
}
# for example table "tbl.tbl" has one integer column, we need to define the signature of input data frame
.sig <- list("num" = "int")
x <- db.gpapply(t, output.name = NULL, FUN = fn.function_plus_one, output.signature = .sig, clear.existing = TRUE, case.sensitive = TRUE, language = "plcontainer")
# comparing with gpapply, gptapply allows you to add an extra argument "INDEX"
# if the table is indexed by column num, we can call gptapply in this way:
.index <- 'num'
y <- db.gptapply(t, INDEX = .index, output.name = NULL, FUN = fn.function_plus_one, output.signature = .sig, clear.existing = TRUE, case.sensitive = TRUE, language = "plcontainer")
For more information about gpapply and gptapply, please refer to detailed document:
TBA