Rule-Based Conformance Checking

2018-10-01

The goal of processcheckR is to support rule-based conformance checking. Currently the following declarative rules can be checked:

Cardinality rules: * contains: activity occurs n times or more * contains_exactly: activity occurs exactly n times * contains_between: activity occures between min and max number of times * absent: activity does not occur more than n - 1 times Ordering rules: * starts: case starts with activity * ends: case ends with activity * succession: if activity A happens, B should happen after. If B happens, A should have happened before. * response: if activity A happens, B should happen after * precedence: if activity B happens, A should have happend before * responded_existence: if activity A happens, B should also (have) happen(ed) (i.e. before or after A) Exclusiveness: * and: two activities always exist together * xor: two activities are not allowed to exist together

Rules can be checked using the check_rule function (see example below). It will create a new logical variable to indicate for which cases the rule holds. The name of the variable can be configured using the label argument in check_rule.

Installation

You can install processcheckR from github with:

# install.packages("devtools")
devtools::install_github("gertjanssenswillen/processcheckR")

Example

library(bupaR)
#>
#> Attaching package: 'bupaR'
#> The following object is masked from 'package:stats':
#>
#>     filter
#> The following object is masked from 'package:utils':
#>
#>     timestamp
library(processcheckR)
#>
#> Attaching package: 'processcheckR'
#> The following object is masked from 'package:base':
#>
#>     xor
sepsis %>%
# check if cases starts with "ER Registration"
check_rule(starts("ER Registration"), label = "r1") %>%
# check if activities "CRP" and "LacticAcid" occur together
check_rule(and("CRP","LacticAcid"), label = "r2") %>%
group_by(r1, r2) %>%
n_cases()
#> # A tibble: 4 x 3
#> # Groups:   r1 [?]
#>   r1    r2    n_cases
#>   <lgl> <lgl>   <int>
#> 1 FALSE FALSE      10
#> 2 FALSE TRUE       45
#> 3 TRUE  FALSE     137
#> 4 TRUE  TRUE      858