db: Read from a Database (currently only PostgreSQL) connection.

In adjust/rport: Connection management and SQL parallelisation for R analytics on big database clusters.

Description

Read from a Database (currently only PostgreSQL) connection.

Usage

db(con.names, sql, params = c(), cores = 4, outfile = "")

Arguments

con.names	a vector of connection names as defined in database.yml (or using custom connection definitions). db parallelizes if multiple connections or queries are given. If more than one connection names is given then the same query is performed on all connections in parallel. This is particularly useful for analytical queries on sharded setup. For example: shards <- paste('shard', 1:16, sep='') db(shards, 'select count(*) from events')) will run in parallel on all 16 shards. If more than one SQL queries is given, then each of them are run in parallel on the single DB connection. If the same length of connections and the same length of SQL queries is given, they are parallelized in pairs. See https://github.com/adjust/rport/ for examples.
params	binds SQL parameters to the SQL query using parameter binding. The PostgreSQL R driver takes care for the quoting. Parameter binding is very important against SQL injection. For example, to get id=123: db(shards, 'select count(*) from events where id = $1', 123)
cores	determines the size of the parallel cluster for parallel queries.
outfile	the outfile variable passed on to makeCluster

See Also

db.connection, db.disconnect, list.connections, reload.db.config, register.connections

adjust/rport documentation built on May 10, 2019, 5:55 a.m.