internals: Internal functions and methods

Description

These functions and methods are used internally and should usually not be called from outside the package.

Usage

begin(x, ...)
  nextPairs(x, n = 10000, ...)
  clear(x, ...)
  blockfldfun(blockfld, phoneticFld, phoneticFun, coln)
  getSQLStatement(data1, data2 = data1, con, type, blockFld, excludeFld,
    strcmpFld, strcmpFun, phoneticFld, phoneticFun)
  init_sqlite_extensions(db)
  .allows_extensions(db)
  .lib_path()
  getPatternCounts(x, n=10000, cutoff=1, withProgressBar = (sink.number()==0))
  getMatchCount(object)
  getNonMatchCount(object)
  getNACount(object)
  getColumnNames(object, withExcluded = FALSE)
  getThresholds(W, M, U, my, ny)
  countpattern(x, matching = FALSE)
  hasWeights(object)

Arguments

x, object: Depends on the function, usually a "RecLinkData", "RLBigData" or "RLResult". A matrix of binary observations for countpattern.
n: Maximum number of comparison patterns to fetch.
blockfld: Blocking definition as in RLBigData*.
phoneticFld: Indices of attributes for which phonetic code should be used.
phoneticFun: Name of phonetic function as in RLBigData*.
data1, data2: Data frame with records.
type: "linkage" or "deduplication", the type of linkage.
excludeFld: Indices or names of fields to exclude from comparison.
strcmpFld: Indices of attributes for which string comparison should be used.
strcmpFun: Name of string comparison function as in RLBigData*.
con: A SQLite database connection.
coln: Column names of records.
db: Database connection.
cutoff: Threshold from which string comparison values are rounded to 1.
withProgressBar: Whether to display a progress bar.
withExcluded: Logical. Whether to output names of excluded columns
W,M,U: Linkage weights, m- and u-probabilities. See emWeights for details.
my,ny: Error bounds. See emClassify for details.
...: Optional arguments.
matching: If TRUE an additional vector is returned which stores which row belongs to which pattern.

Author

Andreas Borg, Murat Sariyar

Details

begin: Sends a query to the database of an "RLBigData" object which creates comparison patterns.
nextPairs: Retrieves the next n comparison patterns.
clear: Closes the result set opened by begin.
blockfldfun: Constructs the part of SQL statement which implements blocking.
blockfldfun: Constructs the part of SQL statement which implements blocking.
getSQLStatement: Constructs SQL statement for retrieving comparison patterns.
init_sqlite_extensions: Load string comparison and phonetic functions into database.
.allows_extensions: Tests whether SQLite engine supports extension functions.
.lib_path: Gets path of shared library of package.
getPatternCounts: Counts binary patterns, implemented through countpattern.
getMatchCount: Gets number of matches from a "RLBigData" object.
getNACount: Gets number of pairs with unknown matching status from a "RLBigData" object.
getDbFile: Returns path of database file for a data or result object.
getColumnNames: Returns column names of a "RLBigData" object as a character vector.
getThresholds: Calculates upper and lower classification based on error bounds.
countpattern: Modified version of countpattern in package e1071.
hasWeights: Determines whether a data object has weights.