db_copy_to: S3 implementation of 'db_copy_to' for Athena

db_copy_toR Documentation

S3 implementation of db_copy_to for Athena

Description

This is an Athena method for dbplyr function db_copy_to to create an Athena table from a data.frame.

Usage

db_copy_to.AthenaConnection(
  con,
  table,
  values,
  overwrite = FALSE,
  append = FALSE,
  types = NULL,
  partition = NULL,
  s3_location = NULL,
  file_type = c("csv", "tsv", "parquet"),
  compress = FALSE,
  max_batch = Inf,
  ...
)

Arguments

con

A dbConnect object, as returned by dbConnect()

table

A character string specifying a table name. Names will be automatically quoted so you can use any sequence of characters, not just any valid bare table name.

values

A data.frame to write to the database.

overwrite

Allows overwriting the destination table. Cannot be TRUE if append is also TRUE.

append

Allow appending to the destination table. Cannot be TRUE if overwrite is also TRUE. Existing Athena DDL file type will be retained and used when uploading data to AWS Athena. If parameter file.type doesn't match AWS Athena DDL file type a warning message will be created notifying user and RAthena will use the file type for the Athena DDL.

types

Additional field types used to override derived types.

partition

Partition Athena table (needs to be a named list or vector) for example: c(var1 = "2019-20-13")

s3_location

s3 bucket to store Athena table, must be set as a s3 uri for example ("s3://mybucket/data/")

file_type

What file type to store data.frame on s3, RAthena currently supports ["tsv", "csv", "parquet"]. Default delimited file type is "tsv", in previous versions of RAthena (=< 1.4.0) file type "csv" was used as default. The reason for the change is that columns containing Array/JSON format cannot be written to Athena due to the separating value ",". This would cause issues with AWS Athena. Note: "parquet" format is supported by the arrow package and it will need to be installed to utilise the "parquet" format.

compress

FALSE | TRUE To determine if to compress file.type. If file type is ["csv", "tsv"] then "gzip" compression is used, for file type "parquet" "snappy" compression is used.

max_batch

Split the data frame by max number of rows i.e. 100,000 so that multiple files can be uploaded into AWS S3. By default when compression is set to TRUE and file.type is "csv" or "tsv" max.batch will split data.frame into 20 batches. This is to help the performance of AWS Athena when working with files compressed in "gzip" format. max.batch will not split the data.frame when loading file in parquet format. For more information please go to link

...

other parameters currently not supported in RAthena

Value

db_copy_to returns table name

See Also

AthenaWriteTables

Examples

## Not run: 
# Note: 
# - Require AWS Account to run below example.
# - Different connection methods can be used please see `RAthena::dbConnect` documnentation

library(DBI)
library(dplyr)

# Demo connection to Athena using profile name 
con <- dbConnect(RAthena::athena())

# List existing tables in Athena
dbListTables(con)

# Write data.frame to Athena table
copy_to(con, mtcars,
        s3_location = "s3://mybucket/data/")
             
# Checking if uploaded table exists in Athena
dbExistsTable(con, "mtcars")

# Write Athena table from tbl_sql
athena_mtcars <- tbl(con, "mtcars")
mtcars_filter <- athena_mtcars %>% filter(gear >=4)

copy_to(con, mtcars_filter)

# Checking if uploaded table exists in Athena
dbExistsTable(con, "mtcars_filter") 

# Disconnect from Athena
dbDisconnect(con)

## End(Not run)

RAthena documentation built on Dec. 28, 2022, 1:19 a.m.