README.md

In lanalabs/LanaR: Process Data from a Lana model

LanaR: API for LANA Process Mining

This package provides an R API for LANA Process Mining.

Attention: This package is still in alpha state. Functions and parameters may be renamed and changed at any time.

Setup

In order to use lanar you need to manually install the packages xml2, plyr using install.packages("xml2", "plyr").

Usage

First upload your event logs to your Lana software. Then enter the log name as a parameter to the different functions as they will use the last uploaded log of that name. Look also at the examples.

Download Sample Event Log

downloads an example event log to your working directory

downloadExample()

A list with all your uploaded logs

getLogs()

| id | owner | name | date | timezone | | ------------- | ------------- | ------------- | ------------- | ------------- | | 546 | 34 | Incident_withImpactAttributes.csv| 2018-09-28T13:19:26.175Z | Etc/GMT | | 547 | 34 | SalesProcess.csv | 2018-09-13T13:27:50.414Z | Europe/Berlin |

Filter and performance / followers statistics

filter(logName)

Example

discoveredModel("Incident_withImpactAttributes.csv")

Activity Performance Statistics

gives statistics about activities such as minimum/maximum duration, average/median/total duration, standard deviation and frequency.

activityPerformance(logName)

Example

activityPerformance("Incident_withImpactAttributes.csv")

| .id | frequency | totalDuration | minDuration | maxDuration | avgDuration | standardDeviation| median | | ------------- | ------------- | ------------- | ------------- | ------------- | ------------- | ------------- | ------------- | | Initial diagnosis | 2337 | 1270500000 | 0 | 1320000 | 543645.7 | 238600.5 | 540000 | | Functional escalation | 851 | 254640000 | 0 | 660000 | 299224.4 | 118595.1 | 300000 | | Incident closure | 2000 | 1582380000 | 0 | 2280000 | 791190.0 | 361650.6 | 780000 | | Incident classification | 2000 | 497580000 | 0 | 900000 | 248790.0 | 166740.6 | 240000 | | Investigation and diagnosis | 851 | 16600800000 | 0 | 60240000 | 19507403.1 | 12030401.3 | 19020000 | | Incident logging | 2000 | 1079040000 | 0 | 2040000 | 539520.0 | 389029.4 | 480000 | | Resolution and recovery | 1635 | 2944380000 | 0 | 6240000 | 1800844.0 | 1111854.5 | 1800000 |

Aggregation

Aggregation with different dimensions such as: time, attribute, frequency, average duration, median duration, total duration.

aggregate(logName, xDimension, yDimension)

Aggregation Dimensions

| Time Dimension | what it does | | ------------- | ------------- | | byTime=byMonth | aggregate by month | | byTime=dayOfWeek | aggregate by day of week | | byTime=byHour | aggregate by day of week | | | |

| Aggregation Dimension | what it does | | ------------- | ------------- | | frequency | counts the cases given time dimension | | avgDuration | outputs the average duration given time dimension | | medianDuration | outputs the median duration given time dimension | | totalDuration | outputs the total duration given time dimension |

Advanced Optional Dimensions

If you are more experienced with Lana, you can insert any kind of filter that you use in Lana into your aggregation or discovered model functions, by simply adding a tracefilter sequence.

TraceFilterSequence:

Variant Filter

[{"max":max variants,"min":min variants,"type":"variantSliderFilter"}]

Attribute Filter

[{"type":"attributeFilter","attributeName":"attribute name","values":["first value","second value"],"inverted":true or false}]

Case Duration Filter

[{"type":"traceDurationFilter","minTraceDuration":min duration in miliseconds,"maxTraceDuration":max duration in miliseconds}]

Timespan Filter

[{"from":start time in miliseconds,"to":ènd time in miliseconds,"startInRange":true,"endInRange":true,"inverted":false,"type":"timeRangeFilter"}]

Activity Filter

[{"type":"activityFilter","activity":"activity1","inverted":true or false},{"type":"activityFilter","activity":"activity 2","inverted":true or false}]

Follower Filter

[{"direct":true or false,"pre":"from","succ":"to","useDuration":true or false,"inverted":true or false,"type":"followerFilter"}]

Endpoint Filter

[{"type":"endpointFilter","activities":["activity1","activity2"],"inverted":true or false}]

Hide Activities Filter

[{"type":"endpointFilter","activities":["activity1","activity2"],"inverted":true or false}]

Examples

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=byMonth", yDimension = "frequency")

| byTime=byMonth | frequency | Case Count | | ------------- | ------------- | ------------- | | Jan 2016 | 2000 | 2000 |

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=dayOfWeek", yDimension = "avgDuration")

| byTime=dayOfWeek | avgDuration | Case Count | | ------------- | ------------- | ------------- | | Monday | 42516060 | 2000 |

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=byHour", yDimension = "medianDuration")

| byTime=byHour | medianDuration | Case Count | | ------------- | ------------- | ------------- | | 09 | 42516060 | 2000 | | 10 | 42516060 | 2000 | | 11 | 42516060 | 2000 | | 12 | 42516060 | 2000 | | 13 | 42516060 | 2000 | | 14 | 42516060 | 2000 | | 15 | 42516060 | 2000 | | 16 | 42516060 | 2000 |

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=byMonth", yDimension = "totalDuration")

| byTime=byMonth | totalDuration | Case Count | | ------------- | ------------- | ------------- | | Jan 2016 | 85032120000 | 2000 |

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=byMonth", yDimension = "totalDuration", traceFilterSequence = [{"max":```max variants```,"min":```min variants```,"type":"variantSliderFilter"}])

aggregate("Incident_withImpactAttributes.csv", xDimension = "byTime=byMonth", yDimension = "totalDuration", traceFilterSequence = [{"max":```max variants```,"min":```min variants```,"type":"variantSliderFilter"}, {"type":"attributeFilter","attributeName":"```attribute name```","values":["```first value```","```second value```"],"inverted":```false or true```}])

lanalabs/LanaR documentation built on Sept. 28, 2021, 8:55 a.m.