Skip to contents

Create a Training-Time Blocking Rule

Source: R/il_block_on.R
block_on.Rd

Creates a blocking rule for use inside training verbs such as il_estimate_em() and il_estimate_prior(). This is distinct from il_block_on(), which adds prediction-time blocking to a specification. The returned object describes how to partition pairs during training.

Usage

block_on(..., .where = NULL, .transform = NULL, .explode = NULL)

Arguments

...

Column names (bare or column ~ transform formulas). See il_block_on() for details on the formula syntax. Columns are AND-ed within a single block_on() call.

.where

An optional raw SQL string for non-equality blocking conditions (e.g., "levenshtein(l.dob, r.dob) <= 1"). When supplied alongside column names, the column equalities and the SQL condition are AND-ed together.

.transform

An optional transform applied to every column that does not already have a formula transform. See il_block_on() for details.

.explode

An optional character vector of array column names to unnest before blocking. See il_block_on() for details.

Value

A blocking-rule object for use in training verbs.

Examples

block_on(first_name, surname)
#> $columns
#> [1] "first_name" "surname"   
#> 
#> $where
#> NULL
#> 
#> $transform
#> NULL
#> 
#> $explode
#> NULL
#> 
#> attr(,"class")
#> [1] "il_blocking_rule"

# Fuzzy SQL conditions
block_on(first_name, .where = 'levenshtein(l.dob, r.dob) <= 1')
#> $columns
#> [1] "first_name"
#> 
#> $where
#> [1] "levenshtein(l.dob, r.dob) <= 1"
#> 
#> $transform
#> NULL
#> 
#> $explode
#> NULL
#> 
#> attr(,"class")
#> [1] "il_blocking_rule"

# Phonetic blocking
block_on(first_name, .transform = il_soundex)
#> $columns
#> [1] "first_name"
#> 
#> $where
#> NULL
#> 
#> $transform
#> function (x) 
#> {
#>     vapply(x, soundex_one, character(1), USE.NAMES = FALSE)
#> }
#> <bytecode: 0x55ee660d0a60>
#> <environment: namespace:irelink>
#> 
#> $explode
#> NULL
#> 
#> attr(,"class")
#> [1] "il_blocking_rule"

# Per-column substring blocking
block_on(first_name ~ il_substr(1, 3), surname ~ il_substr(1, 4))
#> $columns
#> [1] "first_name" "surname"   
#> 
#> $where
#> NULL
#> 
#> $transform
#> $transform$first_name
#> function (x) 
#> substr(x, start, start + length - 1L)
#> <bytecode: 0x55ee6606e100>
#> <environment: 0x55ee6606a960>
#> attr(,"transform_type")
#> [1] "il_substr"
#> attr(,"params")
#> attr(,"params")$start
#> [1] 1
#> 
#> attr(,"params")$length
#> [1] 3
#> 
#> attr(,"class")
#> [1] "il_column_transform" "function"           
#> 
#> $transform$surname
#> function (x) 
#> substr(x, start, start + length - 1L)
#> <bytecode: 0x55ee6606e100>
#> <environment: 0x55ee66069ca0>
#> attr(,"transform_type")
#> [1] "il_substr"
#> attr(,"params")
#> attr(,"params")$start
#> [1] 1
#> 
#> attr(,"params")$length
#> [1] 4
#> 
#> attr(,"class")
#> [1] "il_column_transform" "function"           
#> 
#> 
#> $explode
#> NULL
#> 
#> attr(,"class")
#> [1] "il_blocking_rule"