| Type: | Package |
| Title: | Recursive Mining for Frequent Pattern and Confident Association Rules |
| Version: | 1.0 |
| Date: | 2020-11-14 |
| Author: | Louis Raimbault [aut,cre], A.Mokkadem [aut], M.Pelletier [aut] |
| Maintainer: | Louis Raimbault <Louis.Raimbault@icloud.com> |
| Description: | Provides functions allowing the user to recursively extract frequent patterns and confident rules according to indicators of minimal support and minimal confidence. These functions are described in "Recursive Association Rule Mining" Abdelkader Mokkadem, Mariane Pelletier, Louis Raimbault (2020) <doi:10.48550/arXiv.2011.14195>. |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| Imports: | Rcpp (≥ 1.0.5) |
| LinkingTo: | Rcpp |
| URL: | https://github.com/LouisRaimbault/RecAssoRules-R |
| BugReports: | https://github.com/LouisRaimbault/RecAssoRules-R |
| NeedsCompilation: | yes |
| RoxygenNote: | 7.1.1 |
| Packaged: | 2020-12-01 23:15:02 UTC; lraimbault |
| Repository: | CRAN |
| Date/Publication: | 2020-12-04 10:30:02 UTC |
Recursive mining for frequent itemSet and confident association rules.
Description
This first version makes it possible to carry out the usual applications of association rule Mining.This by using recursive algorithms. All of the functions are using the C++ language. For more usage security, it mainly uses the std and stl libraries, reducing only very slightly the efficiency of the algorithms. It will be updated soon with functions allowing it to be applied on non-fixed databases, as well as a more practical interface.
Transacfruits Data Set
Description
The Transacfruits dataset is a synthetic database. It thus represents fictitious transactions, simulating fruit purchases, among a random list.
Usage
data("Transacfruits")
Details
Transacfruits is made up of 1 columns and 100 rows. There are therefore 100 transactions. For each of them one or more fruits were purchased from a list of 15 fruits. It was generated by binomial simulation, for all transactions, there was for each fruit a fixed probability p of being bought.
Mining for frequent patterns and confident rules.
Description
Returns a list with two objets : freqset a R DataFrame with 3 columns , the set-names , their support and relative support and rulesset a R Dataframe with 3 colums , the antecedant names , the consequent names and their confidence. When lauching the function, the BinaryMatrix is ordered in ascendant colsums order. For each column, the frequent itemSets and confident Rules containing the current tested item are extracted with the ARM traversal function.
Usage
prefrecrules (Bitmax , varnames , relativeSup , Minconf )
Arguments
Bitmax |
A binary DataFrame. |
varnames |
A vector of character or a Matrix character. |
relativeSup |
The wanted relative min Support value. |
Minconf |
The wanted min confident value. |
Examples
data("Transacfruits")
## Import and build Transafruits as binary matrix.
binaryfruits = transactiontoBitmax (as.matrix(Transacfruits),",")
outsetrules = prefrecrules (binaryfruits, names(binaryfruits),0.02, 0.4)
## Mine frequents itemsets and confidentes rules with :
## relative min support of 0.02 and min confidence of 0.4.
Mining for frequent patterns.
Description
Provides a R DataFrame with 3 columns , the set-names , their support and relative support. When lauching the function, The BinaryMatrix is ordered in ascendant colsums order. For each column, the frequent itemSets containing the current tested item are extracted with the traversal function.
Usage
prefrecset (Bitmax , varnames , relativeSup )
Arguments
Bitmax |
A binary DataFrame |
varnames |
A vector of character or a Matrix character. |
relativeSup |
The wanted relative min Support value. |
Examples
data("Transacfruits")
## Import and build Transafruits as binary matrix.
binaryfruits = transactiontoBitmax (as.matrix(Transacfruits),",")
itemsets = prefrecset (binaryfruits, names(binaryfruits), 0.02)
## Mine frequents itemSet from Transacfruits with a relative min support of 0.02.
Creation of a Binary DataFRame from transactions type
Description
Returns the corresponding Binary R DataFrame. If you imported your transaction as R DataFrame, please use the R function as.matrix(yourdataframe).
Usage
transactiontoBitmax(transac , deli)
Arguments
transac |
a character vector or matrix. |
deli |
The character used to delimit the objects present in a transaction. |
Examples
data("Transacfruits")
binaryfruits = transactiontoBitmax (as.matrix(Transacfruits),",")
## Import transaction data Transacfruits and create the correspondant binary matrix.