The goal of pmap is to provide functionality of generating process map from given event logs.
This is a basic example which shows you how to solve a common problem:
## basic example code
library(pmap)
library(dplyr)
# Generate simulated eventlog
> eventlog <- generate_eventlog(
size_of_eventlog = 10000,
number_of_customers = 2000,
event_catalogs = c("campaign", "sale"),
event_catalogs_size = c(10, 4))
# Check eventlog data frame structure
> head(eventlog)
timestamp customer_id event_name event_type
1 2017-01-01 00:08:35 Customer 1072 Event 6 (campaign) campaign
2 2017-01-01 00:14:42 Customer 1979 Event 1 (campaign) campaign
3 2017-01-01 00:55:58 Customer 32 Event 3 (campaign) campaign
4 2017-01-01 01:04:01 Customer 1877 Event 8 (campaign) campaign
5 2017-01-01 01:47:37 Customer 833 Event 7 (campaign) campaign
6 2017-01-01 03:34:39 Customer 1119 Event 4 (campaign) campaign
> str(eventlog)
'data.frame': 10000 obs. of 4 variables:
$ timestamp : POSIXct, format: "2017-01-01 00:08:35" "2017-01-01 00:14:42" ...
$ customer_id: chr "Customer 1072" "Customer 1979" "Customer 32" "Customer 1877" ...
$ event_name : chr "Event 6 (campaign)" "Event 1 (campaign)" "Event 3 (campaign)" "Event 8 (campaign)" ...
$ event_type : chr "campaign" "campaign" "campaign" "campaign" ...
# Create process map
> p <- create_pmap(eventlog, target_types = c("sale"))
# Render the process map
> print(render_pmap(p))The result will be a bit messy.
Let’s prune the process map.
# Prune the process map
> p <- p %>% prune_edges(0.5) %>% prune_nodes(0.5)
# Render the pruned process map
> print(render_pmap(p))