Filtering in relational data models

Katharina Brunner


The {dm} package offers functions to work with relational data models in R.

This document introduces you to filtering functions, and shows how to apply them to the data that is separated into multiple tables.

Our example data is drawn from the {nycflights13} package that contains five inter-linked tables.

First, we will load the packages that we need:

#> Error in library(nycflights13): there is no package called 'nycflights13'

Data: nycflights13

To explore filtering with {dm}, we’ll use the {nycflights13} data with its flights, planes, airlines, airports and weather tables.

This dataset contains information about the 336 776 flights that departed from New York City in 2013, with 3322 different planes and 1458 airports involved. The data comes from the US Bureau of Transportation Statistics, and is documented in ?nycflights13::flights.

To start with our exploration, we have to create a dm object from the {nycflights13} data. The built-in dm::dm_nycflights13() function takes care of this.

By default it only uses a subset of the complete data though: only the flights on the 10th of each month are considered, reducing the number of rows in the flights table to 11 227.

A data model object contains data from the source tables, and metadata about the tables.

If you would like to create a dm object from tables other than the example data, you can use the new_dm(), dm() or as_dm() functions. See vignette("howto-dm-df") for details.

dm <- dm_nycflights13()

The console output of the ’dm` object shows its data and metadata, and is colored for clarity:

#> ── Metadata ───────────────────────────────────────────────────────────────
#> Tables: `airlines`, `airports`, `flights`, `planes`, `weather`
#> Columns: 53
#> Primary keys: 4
#> Foreign keys: 4

Now we know that there are five tables in our dm object. But how are they connected? These relations are best displayed as a visualization of the entity-relationship model:

#> Error in loadNamespace(x): there is no package called 'DiagrammeRsvg'

You can look at a single table with tbl. To print the airports table, call

tbl(dm, "airports")
#> Warning: `` was deprecated in dm 0.2.0.
#> Use `dm[[table_name]]` instead to access a specific table.
#> # A tibble: 86 x 8
#>    faa   name                     lat    lon   alt    tz dst   tzone       
#>    <chr> <chr>                  <dbl>  <dbl> <dbl> <dbl> <chr> <chr>       
#>  1 ALB   Albany Intl             42.7  -73.8   285    -5 A     America/New…
#>  2 ATL   Hartsfield Jackson At…  33.6  -84.4  1026    -5 A     America/New…
#>  3 AUS   Austin Bergstrom Intl   30.2  -97.7   542    -6 A     America/Chi…
#>  4 BDL   Bradley Intl            41.9  -72.7   173    -5 A     America/New…
#>  5 BHM   Birmingham Intl         33.6  -86.8   644    -6 A     America/Chi…
#>  6 BNA   Nashville Intl          36.1  -86.7   599    -6 A     America/Chi…
#>  7 BOS   General Edward Lawren…  42.4  -71.0    19    -5 A     America/New…
#>  8 BTV   Burlington Intl         44.5  -73.2   335    -5 A     America/New…
#>  9 BUF   Buffalo Niagara Intl    42.9  -78.7   724    -5 A     America/New…
#> 10 BUR   Bob Hope                34.2 -118.    778    -8 A     America/Los…
#> # … with 76 more rows

Filtering a dm object

dm_filter() allows you to select a subset of a dm object.

How it works

Filtering a dm object is not that different from filtering a dataframe or tibble with dplyr::filter().

The corresponding {dm} function is dm::dm_filter(). With this function one or more filtering conditions can be set for one of the tables of the dm object. These conditions are immediately evaluated for their respective tables and in addition they are stored in the dm. There are two ways in which a condition on one table could affect another table:

  1. Calling dm_apply_filters_to_tbl(), i.e. requesting a table from a dm after the filter conditions are applied. In this case, all tables that are connected to the requested table and have stored filter conditions associated with them are taken into account in the following way:
    • filtering semi-joins are successively performed along the paths from each of the filtered tables to the requested table, each join reducing the left-hand side tables of the joins to only those of their rows with key values that have corresponding values in key columns of the right-hand side tables of the join.
    • eventually the requested table is returned, containing only the the remaining rows after the filtering joins
  2. Calling dm_apply_filters() or compute() methods for dm objects on a dm: this results in a new dm that contains the same tables as before. Each table is the result of a dm_apply_filters_to_tbl() call, meaning that the effects of the filter conditions on each of the tables are taken into account.

Currently, this only works if the graph induced by the foreign key relations is cycle free. Fortunately, this is the default for dm_nycflights13().

Keep in mind that several {dm} functions will refuse to work when unevaluated filter conditions exist, such as, e.g. dm_enum_fk_candidates(), dm_enum_pk_candidates(), dm_select_tbl(), dm_rename_tbl(), dm_select(), dm_rename() and dm_nrow(). In these cases consider applying the filters with dm_apply_filters() first.

Filtering Examples

Let’s see filtering in action:

We only want the data that is related to John F. Kennedy International Airport.

filtered_dm <-
  dm %>%
  dm_filter(airports, name == "John F Kennedy Intl")
#> ── Metadata ───────────────────────────────────────────────────────────────
#> Tables: `airlines`, `airports`, `flights`, `planes`, `weather`
#> Columns: 53
#> Primary keys: 4
#> Foreign keys: 4
#> ── Filters ────────────────────────────────────────────────────────────────
#> airports: name == "John F Kennedy Intl"

The filter expression is listed in the print output.

You can get the numbers of rows of each table with dm_nrow(). Before doing that, you will need to apply the filters using dm_apply_filters():

rows_per_table <-
  filtered_dm %>%
  dm_apply_filters() %>%
#> airlines airports  flights   planes  weather 
#>       10        1      602      336       38
#> [1] 987
sum_nrow <- sum(dm_nrow(dm))
sum_nrow_filtered <- sum(dm_nrow(dm_apply_filters(filtered_dm)))

The total number of rows in the dm drops from 2 951 to 987 (the only unaffected table is the disconnected weather table).

Next example:

Get a dm object containing data for flights from New York to the Dulles International Airport in Washington D.C., abbreviated with IAD.

dm %>%
  dm_filter(flights, dest == "IAD") %>%
  dm_apply_filters() %>%
#> airlines airports  flights   planes  weather 
#>        4        3       32       28       30

Chaining multiple filters on different tables is also supported.

An example:

Get all flights from Delta Air Lines which didn’t depart from John F. Kennedy International Airport in May 2013.

dm_delta_may <-
  dm %>%
  dm_filter(airlines, name == "Delta Air Lines Inc.") %>%
  dm_filter(airports, name != "John F Kennedy Intl") %>%
  dm_filter(flights, month == 5)
#> ── Metadata ───────────────────────────────────────────────────────────────
#> Tables: `airlines`, `airports`, `flights`, `planes`, `weather`
#> Columns: 53
#> Primary keys: 4
#> Foreign keys: 4
#> ── Filters ────────────────────────────────────────────────────────────────
#> airlines: name == "Delta Air Lines Inc."
#> airports: name != "John F Kennedy Intl"
#> flights: month == 5
dm_delta_may %>%
  dm_apply_filters() %>%
#> airlines airports  flights   planes  weather 
#>        0        0        0        0        0

You can inspect the filtered tables with dm_apply_filters_to_tbl().

In the airlines table, Delta is the only remaining carrier:

dm_delta_may %>%
#> # A tibble: 0 x 2
#> # … with 2 variables: carrier <chr>, name <chr>

Which planes were used to service these flights?

dm_delta_may %>%
#> # A tibble: 0 x 9
#> # … with 9 variables: tailnum <chr>, year <int>, type <chr>,
#> #   manufacturer <chr>, model <chr>, engines <int>, seats <int>,
#> #   speed <int>, engine <chr>

And indeed, all included flights departed in May (month == 5):

dm_delta_may %>%
#> # A tibble: 0 x 19
#> # … with 19 variables: year <int>, month <int>, day <int>, dep_time <int>,
#> #   sched_dep_time <int>, dep_delay <dbl>, arr_time <int>,
#> #   sched_arr_time <int>, arr_delay <dbl>, carrier <chr>, flight <int>,
#> #   tailnum <chr>, origin <chr>, dest <chr>, air_time <dbl>,
#> #   distance <dbl>, hour <dbl>, minute <dbl>, time_hour <dttm>

For comparison, let’s review the equivalent manual query for flights in dplyr syntax:

airlines_filtered <- filter(airlines, name == "Delta Air Lines Inc.")
#> Error in filter(airlines, name == "Delta Air Lines Inc."): object 'airlines' not found
airports_filtered <- filter(airports, name != "John F Kennedy Intl")
#> Error in filter(airports, name != "John F Kennedy Intl"): object 'airports' not found
flights %>%
  semi_join(airlines_filtered, by = "carrier") %>%
  semi_join(airports_filtered, by = c("origin" = "faa")) %>%
  filter(month == 5)
#> Error in semi_join(., airlines_filtered, by = "carrier"): object 'flights' not found

The {dm} code is leaner because the foreign key information is encoded in the object.

SQL statements behind filtering a dm object on a database

{dm} is meant to work with relational data models, locally as well as on databases. In your project, the data is probably not stored locally but in a remote relational database that can be queried with SQL statements.

You can check the queries by using sql_render() from the {dbplyr} package.


Print the SQL statements for getting all flights from Delta Air Lines, which did not depart from John F. Kennedy International Airport in May 2013, with the data stored in a sqlite database.

To show the SQL query behind a dm_filter(), we copy the flights, airlines and airports tables from the nyflights13 dataset to a temporary in-memory database using the built-in function copy_dm_to() and dbplyr::src_memdb.

Then we filter the data, and print the corresponding SQL statement with dbplyr::sql_render().

dm %>%
  dm_select_tbl(flights, airlines, airports) %>%
  dm_filter(flights, month == 5) %>%
  copy_dm_to(dbplyr::src_memdb(), .) %>%
  dm_filter(airlines, name == "Delta Air Lines Inc.") %>%
  dm_filter(airports, name != "John F Kennedy Intl") %>%
  dm_apply_filters() %>%
  dm_get_tables() %>%
#> Error in (function (cond) : error in evaluating the argument 'drv' in selecting a method for function 'dbConnect': there is no package called 'RSQLite'

Further reading: {dm}’s function for copying data from and to databases.