The downside of refusals
In the last exercise, we looked at the cases in which we didn't restart the procedure or didn't decline candidates. In these cases, it seems that the vacancies were filled slightly faster. But what is the impact when the candidate refuses our proposals?
filter_precedence()
can be used to select cases based on specific orders between activities by providing antecedent and consequent activities.
Its usage is
dataset %>%
filter_precedence(
antecedents = "activity that came before",
consequents = "activity that came after",
precedence_type = "directly or eventually following?"
filter_method = "all or none?"
)
This exercise is part of the course
Business Process Analytics in R
Exercise instructions
- Filter for the cases in which
"Receive Response"
is never followed directly by"Review Non Acceptance"
. Store it innot_refused
. This can be done by settingprecedence_type
to"directly_follows"
andfilter_method
to"none"
. - The cases in
not_refused
still have a high throughput time. Let's select those cases ofnot_refused
which take longer than 300 days. Store them inworst_cases
. - Use
trace_explorer()
to look at the 80% most frequent different traces in theworst_cases
event log.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Filter by precedence with activites not directly following
not_refused <- vacancies %>%
___(
antecedents = ___,
consequents = ___,
precedence_type = ___,
filter_method = ___
)
# Filter for throughput time greater than 300
worst_cases <- ___ %>%
filter_throughput_time(interval = ___, units = ___)
# Explore the traces with 80% coverage
___