Introduction to dbi.table

Query database tables and views over a DBI connection using data.table’s [i, j, by] syntax, attach database schemas to the search path, and programmatically load database catalogs.

This vignette assumes that you are already fluent with data.table’s syntax and that you know how to open a database connection using the DBI package.

1 Installation

The dbi.table package is hosted on GitHub. Use the following command to install the package.

#install.packages("devtools")
devtools::install_github("kjellpk/dbi.table")

Note: if the install_github function is not found, you will need to first install the devtools package using install.packages.

2 Getting Started

This section uses the sample Chinook Database (included in the package) to demonstrate how to

  1. create a single dbi.table using the dbi.table function,
  2. maniuplate a dbi.table using data.table’s [i, j, by] syntax,
  3. attach a schema to the search path using the dbi.attach function, and
  4. load a database catalog using the dbi.catalog function.

The function chinook.duckdb that returns an open duckdb (DBI) connection to the sample Chinook Database. This connection is a typical DBI connection as returned by DBI::dbConnect that can be used as the conn argument in DBI package functions. Let’s get started by loading the package and opening the connection.

library(data.table)
library(dbi.table)
chinook <- chinook.duckdb()

2.1 Create a Single dbi.table

The dbi.table function takes 2 arguments: a DBI connection, and an Id indentifying a database table or view.

my_album <- dbi.table(chinook, DBI::Id("Album"))

The object my_album is a dbi.table, a data structure that represents a SQL query (which we refer to as the dbi.table’s underlying SQL query). The print method displays a preview of the underlying SQL query.

#print(my_album)
my_album
## <chinook_duckdb> Album 
##  AlbumId                                 Title ArtistId
##    <int>                                <char>    <int>
##        1 For Those About To Rock We Salute You        1
##        2                     Balls to the Wall        2
##        3                     Restless and Wild        2
##        4                     Let There Be Rock        1
##        5                              Big Ones        3
##  ---

The preview has a format similar to a data.table with two notable exceptions.

  1. The row numbers are omitted. SQL queries do not necessarily return the result set in a reliable order (even on subsequent evaluations of the same query), and dbi.table does not make any extra effort to order the rows by default. Thus the row numbers are omitted.

  2. Only 5 rows of the dbi.table are displayed (data.table displays the first 5 and the last 5). Again, since the result set does not have a reliable order, it is not possible to say which rows are the first and which are the last. The rows displayed are the first 5 returned by the RDBMS.

The function as.data.table executes the dbi.table’s underlying SQL query and retrieves the result set as a data.table. Pro tip: calling the extracts method ([]) with no arguments is a shortcut for as.data.table.

#as.data.table(my_album)
my_album[]
##      AlbumId                                                        Title ArtistId
##        <int>                                                       <char>    <int>
##   1:       1                        For Those About To Rock We Salute You        1
##   2:       2                                            Balls to the Wall        2
##   3:       3                                            Restless and Wild        2
##   4:       4                                            Let There Be Rock        1
##   5:       5                                                     Big Ones        3
##  ---                                                                              
## 343:     343                                       Respighi:Pines of Rome      226
## 344:     344 Schubert: The Late String Quartets & String Quintet (3 CD's)      272
## 345:     345                                          Monteverdi: L'Orfeo      273
## 346:     346                                        Mozart: Chamber Music      274
## 347:     347           Koyaanisqatsi (Soundtrack from the Motion Picture)      275

Since the result set is instantiated locally as a data.table, the row numbers and the last 5 rows are displayed.

Note: by default, as.data.table (and the empty extracts shortcut) fetch a maximum of 10,000 rows. To override this limit, either set the option dbi_table_max_fetch or call as.data.table and provide the n argument (e.g., n = -1 to fetch the entire result set).

The csql utility displays the query.

csql(my_album)
## SELECT Album.AlbumId AS AlbumId,
##        Album.Title AS Title,
##        Album.ArtistId AS ArtistId
## 
##   FROM Album AS Album
## 
##  LIMIT 10000

The underlying SQL query of a newly created dbi.table selects all the columns from the database table.

2.2 Manipulate a dbi.table using data.table Syntax

This table from data.table’s Introduction to data.table vignette pretty much sums up what dbi.table does.

DT[i, j, by]

##   R:                 i                 j        by
## SQL:  where | order by   select | update  group by

In general, dbi.table should be able to handle basic data.table syntax. SQL translation is done by dbplyr::translate_sql_ which works with a wide variety of R functions. However, complicated expressions (e.g., custom functions in j, nested aggregation functions, most special symbols) do not work.

Best practice is to use dbi.table to subset and wrangle on the database, then data.table to fine tune locally.

The remainder of this section demonstrates how i, j, and by manipulate a dbi.table’s underlying SQL query.

When i is a logical expression of the variables in the dbi.table then it becomes the WHERE clause in the dbi.table’s underlying SQL query.

csql(my_album[AlbumId == ArtistId + 1])
## SELECT Album.AlbumId AS AlbumId,
##        Album.Title AS Title,
##        Album.ArtistId AS ArtistId
## 
##   FROM Album AS Album
## 
##  WHERE Album.AlbumId = (Album.ArtistId + 1)
## 
##  LIMIT 10000

When i is a call to order (or forder), it becomes the ORDER BY clause in the dbi.table’s underlying SQL query.

csql(my_album[order(nchar(Title), -AlbumId)])
## SELECT Album.AlbumId AS AlbumId,
##        Album.Title AS Title,
##        Album.ArtistId AS ArtistId
## 
##   FROM Album AS Album
## 
##  ORDER BY LENGTH(Album.Title), Album.AlbumId DESC
## 
##  LIMIT 10000

When j is a list of expressions of the variables in the dbi.table, then j becomes the SELECT clause in the dbi.table’s underlying SQL query.

csql(my_album[, .(AlbumId, Title)])
## SELECT Album.AlbumId AS AlbumId,
##        Album.Title AS Title
## 
##   FROM Album AS Album
## 
##  LIMIT 10000

When by is a list of expressions of the variables in the dbi.table, then by becomes the GROUP BY clause in the dbi.table’s underlying SQL query.

csql(my_album[, .("# of Albums" = .N), .(ArtistId)])
## SELECT Album.ArtistId AS ArtistId,
##        COUNT(*) AS "# of Albums"
## 
##   FROM Album AS Album
## 
##  GROUP BY Album.ArtistId
## 
##  LIMIT 10000

2.3 Attach a Schema to the Search Path

The dbi.attach function attaches a DBI connection to the search path. That is, dbi.attach creates a dbi.table for each table and each view in the schema associated with the DBI connection, then assigns these dbi.tables to an environment on the search path.

dbi.attach(chinook)

A quick look at the search path shows the database attached in position 2.

head(search(), 3)
## [1] ".GlobalEnv"            "duckdb:chinook_duckdb" "package:dbi.table"

The tables and views in the database schema are queriable as dbi.tables in the attached environment duckdb:chinook_duckdb.

ls("duckdb:chinook_duckdb")
##  [1] "Album"         "Artist"        "Customer"      "Employee"      "Genre"        
##  [6] "Invoice"       "InvoiceLine"   "MediaType"     "Playlist"      "PlaylistTrack"
## [11] "Track"

Note: Attaching a DBI connection is intended for an interactive exploratory analysis of a database (schema). For programatic use cases, see the Load a Database Catalog section.

Merging two dbi.tables results in a SQL join that describes the same result set as the associated data.table merge. That is,

merge(as.data.table(Album), as.data.table(Artist), by = "ArtistId")

and

as.data.table(merge(Album, Artist, by = "ArtistId"))

are the same data.table up to row order.

csql(merge(Album, Artist, by = "ArtistId"))
## SELECT Album.ArtistId AS ArtistId,
##        Album.AlbumId AS AlbumId,
##        Album.Title AS Title,
##        Artist."Name" AS "Name"
## 
##   FROM chinook_duckdb.main.Album AS Album
## 
##  INNER JOIN chinook_duckdb.main.Artist AS Artist
##     ON Album.ArtistId = Artist.ArtistId
## 
##  LIMIT 10000

When a DBI connection is attached to the search path, dbi.attach also loads the schema’s relational meta data (whether this works depends on how the underlying database implments an information schema). In particular, foreign key constraints are used as the default by when merging. In the previous example, the ArtistId column is a foreign key referencing the Album table. For this example, when the by argument is omitted, dbi.table still merges by ArtistId.

csql(merge(Album, Artist))
## SELECT Album.ArtistId AS ArtistId,
##        Album.AlbumId AS AlbumId,
##        Album.Title AS Title,
##        Artist."Name" AS "Name"
## 
##   FROM chinook_duckdb.main.Album AS Album
## 
##  INNER JOIN chinook_duckdb.main.Artist AS Artist
##     ON Album.ArtistId = Artist.ArtistId
## 
##  LIMIT 10000

When the y argument is omitted, dbi.table’s merge uses the foreign key constraints that reference x to determine the y (or ys) to merge with.

csql(merge(Track))
## SELECT Track.MediaTypeId AS MediaTypeId,
##        Track.GenreId AS GenreId,
##        Track.AlbumId AS AlbumId,
##        Track.TrackId AS TrackId,
##        Track."Name" AS "Name",
##        Track.Composer AS Composer,
##        Track."Milliseconds" AS "Milliseconds",
##        Track.Bytes AS Bytes,
##        Track.UnitPrice AS UnitPrice,
##        Album.Title AS "Album.Title",
##        Album.ArtistId AS "Album.ArtistId",
##        Genre."Name" AS "Genre.Name",
##        MediaType."Name" AS "MediaType.Name"
## 
##   FROM chinook_duckdb.main.Track AS Track
## 
##   LEFT OUTER JOIN chinook_duckdb.main.Album AS Album
##     ON Track.AlbumId = Album.AlbumId
## 
##   LEFT OUTER JOIN chinook_duckdb.main.Genre AS Genre
##     ON Track.GenreId = Genre.GenreId
## 
##   LEFT OUTER JOIN chinook_duckdb.main.MediaType AS MediaType
##     ON Track.MediaTypeId = MediaType.MediaTypeId
## 
##  LIMIT 10000

When the optional recursive argument is TRUE, merge.dbi.table recursively merges on each of the just-merged tables. In this example, Track has a foreign key that references Album and Album has a foreign key that references Artist.

csql(merge(Track, recursive = TRUE))
## SELECT Album.ArtistId AS "Album.ArtistId",
##        Track.MediaTypeId AS MediaTypeId,
##        Track.GenreId AS GenreId,
##        Track.AlbumId AS AlbumId,
##        Track.TrackId AS TrackId,
##        Track."Name" AS "Name",
##        Track.Composer AS Composer,
##        Track."Milliseconds" AS "Milliseconds",
##        Track.Bytes AS Bytes,
##        Track.UnitPrice AS UnitPrice,
##        Album.Title AS "Album.Title",
##        Genre."Name" AS "Genre.Name",
##        MediaType."Name" AS "MediaType.Name",
##        Artist."Name" AS "Artist.Name"
## 
##   FROM chinook_duckdb.main.Track AS Track
## 
##   LEFT OUTER JOIN chinook_duckdb.main.Album AS Album
##     ON Track.AlbumId = Album.AlbumId
## 
##   LEFT OUTER JOIN chinook_duckdb.main.Genre AS Genre
##     ON Track.GenreId = Genre.GenreId
## 
##   LEFT OUTER JOIN chinook_duckdb.main.MediaType AS MediaType
##     ON Track.MediaTypeId = MediaType.MediaTypeId
## 
##   LEFT OUTER JOIN chinook_duckdb.main.Artist AS Artist
##     ON Album.ArtistId = Artist.ArtistId
## 
##  LIMIT 10000

2.4 Load a Database Catalog

As a best practice for programatic use, it is better to load the catalog in order to avoid modifying the search path.

catalog <- dbi.catalog(chinook)

Printing the catalog lists its schemas.

catalog
## <Database Catalog> duckdb::chinook_duckdb (2 schemas containing 15 objects) 
## [1] "information_schema" "main"

Individual tables can be accessed using catalog$schema$table syntax.

catalog$main$Album
## <chinook_duckdb> Album 
##  AlbumId                                 Title ArtistId
##    <int>                                <char>    <int>
##        1 For Those About To Rock We Salute You        1
##        2                     Balls to the Wall        2
##        3                     Restless and Wild        2
##        4                     Let There Be Rock        1
##        5                              Big Ones        3
##  ---

When a catalog is loaded, all of its tables have access to the relational data in the information schema.

merge(catalog$main$Album)
## <chinook_duckdb> Album + Artist 
##  ArtistId AlbumId                                 Title Artist.Name
##     <int>   <int>                                <char>      <char>
##         1       1 For Those About To Rock We Salute You       AC/DC
##         2       2                     Balls to the Wall      Accept
##         2       3                     Restless and Wild      Accept
##         1       4                     Let There Be Rock       AC/DC
##         3       5                              Big Ones   Aerosmith
##  ---

3 Scope

This section provides a brief explanation of what the dbi.table package is trying to do.

Suppose that x is a dbi.table and that e is an expression involving x that returns either a dbi.table or a data.table.

x <- dbi.table(chinook, DBI::Id("Album"))
e <- quote(x[, .("# of Albums" = .N), .(ArtistId)])

Since dbi.table’s syntax is a subset of data.table’s syntax, if e can be evaluated successfully (i.e., eval(e) does not throw an error), then e should also be able to be successfully evaluated when x is a data.table. There are thus 2 paths to the final data.table result:

  1. evaluate e then coerce the result using as.data.table, or

  2. coerce x to a data.table then evaluate e.

Path 2 is referred to as the reference implementation and describes the correct answer: the reference result set. The design goal of dbi.table is to get the same result set as the reference result set, up to row order.

result_set <- as.data.table(eval(e))
x <- as.data.table(x)
reference_result_set <- eval(e)
all.equal(reference_result_set, result_set, ignore.row.order = TRUE)
## [1] TRUE

The dbi.table package includes the function reference.test that compares the result set to the reference result set in the more general case where expr (the function’s first argument) is an expression involving 1 or more dbi.tables.

x <- dbi.table(chinook, DBI::Id("Album"))
reference.test({
  x[, .("# of Albums" = .N), .(ArtistId)]
})
## [1] TRUE

This function is used extensively in dbi.table’s unit/regression tests.

4 Cleaning Up

We used the chinook.duckdb function to open a DBI connection at the beginning of this vignette and now it is up to us to close it.

DBI::dbDisconnect(chinook)

However, this leaves our R session in a wonky state. The environment “duckdb:chinook_duckdb” is still attached and there are several dbi.tables in the global environment - all of these dbi.tables are associated with an invalid DBI connection.

#A dbi.table in the duckdb:chinook_duckdb environment
Genre
## Error in `dbSendQuery()`:
## ! rapi_prepare: Invalid connection

The R objects associated with our now-closed DBI connection need to be cleaned up manually (or you could just restart R).

detach("duckdb:chinook_duckdb")
rm(catalog, my_album, x)

4.1 Connection Management

Alternatively, when using either dbi.attach or dbi.catalog, the first arguement can be a zero-argument function that returns an open DBI connection. When dbi.table uses a function to open the DBI connection, then that connection belongs to dbi.table and dbi.table will take care of closing it when it is no longer needed.

dbi.attach(chinook.duckdb)

When dbi.table is managing the connection, then all the user has to do is detach (or delete if a catalog). The DBI connection will be closed when the object is garbage collected.

detach("duckdb:chinook_duckdb")

Further, when dbi.table owns the connection, it is able to reconnect in the event that the connection unexpectedly drops.