coderr helps you write SQL code in a way that presents a clear workflow.
Here is an example that aggregates earthquake information from the built-in
quakes
dataset.:
library(sqldf)
library(coderr)
code_sql_sample("quakes") %>%
code_sql_group_by(aggregator.cols = "avg(mag) AS mag_mean",
group.by.cols = c("stations")) %>%
code_sql_select(select.cols = c("stations", "mag_mean"),
filters = c("stations < 15"),
order.by.cols = c("stations")) %>%
sqldf()
This calculates earthquake magnitude means by station and then presents only stations with station names with a value less than 15.
coderr is used in large scale statistical programming projects where data is stored on remote large scale databases. There's also more to it than that - check out the code and tests to see more.