exploration.livemd

Exploration

Mix.install([
  {:explorer, "~> 0.5.0"}
])

Section

alias Explorer.DataFrame, as: DF

path_to_treatment_data = Path.join([__DIR__, "artificial_patient_treatment.csv"])

dtypes = [
  {"patient", :string},
  {"action", :string},
  {"org:resource", :string},
  {"DateTime", :date}
]

{:ok, df} = DF.from_csv(path_to_treatment_data, dtypes: dtypes)

For the creation of a Directly-Follows-Graph (DFG) there a three parameters by which we can filter

• minimal number of traces for each variant included
• minimal number of events for each activity included
• minimal number of direct successions for each relation included

activities = DF.distinct(df, [" action"])