Quickstart

This is a quick introduction to Pyromorphite. Before proceeding, make sure that Pyromorphite is installed.

Now let’s make sure pyromorphite is imported:

>>> import pyromorphite as pm

Read a Log 📜

Reading in event log files in Pyromorphite is super easy. It supports XES natively and CSV as well as Excel files via pandas.

XES Files

Having previously imported pyromorphite, we can look for a xes file to import. We’ll pick this only repository as it’s full with further examples you can try out

>>> URL = "https://data.4tu.nl/repository/uuid:c1e9137e-2877-410d-a76a-21ce7f97a239/DATA1"
>>> log = pm.read_xes(URL)

CSV and Excel with pandas

Begin by import the pandas module:

>>> import pandas as pd

We’ll try now to get an XES file from a web repository, like this dataset:

>>> URL = "https://data.4tu.nl/repository/uuid:d5ccb355-ca67-480f-8739-289b9b593aaf/DATA"
>>> log = pd.read_csv(URL)

Construct a Bag 🎒

Having parsed a log into a pandas DataFrame we can extract all unique traces together with their frequency in the log as a multiset or bag:

>>> bag = pm.as_bag(log)

We should also consider that not everyone might use the same column naming in their documents:

>>> bag = pm.as_bag(log, case='CI Name (aff)', time='Actual Start', activity='Change Type')

We can therefore specify the column names to be used as:

• case identifiers
• event label
• and timestamp

Mine Your First Model 💃

With our bag of traces in hand we can go onto mining our first model. We’ll go for a process tree. This in not so relevant as most models can be converted between one another. What is relevant, though, is the miner we use. We’ll give the Inductive Miner a try:

>>> ptree = pm.InductiveMiner(bag)