Cluster trend overview
Refers to the average percentage daily increase or decrease in number of incidents created within a cluster.
Example:
You have 100 tickets that are divided into 3 clusters and the look back period is biweekly. The clusters are formed in this manner:
- Cluster 1: 30 tickets
- Cluster 2: 45 tickets
- Cluster 3: 25 tickets
Calculation for each cluster
To determine the average daily trend inside every cluster, the calculation is done in the following manner:
(Difference in the number of tickets created as compared to the previous day/Total tickets created on the previous day) *100
Day | Tickets created per day | Difference in the number of tickets created as compared to the previous day | Percentage increase or decrease w.r.t tickets on previous day |
|---|---|---|---|
Day 1 | 3 | - | - |
Day 2 | 0 | -3 | (-3 / 3 ) * 100 = -100% |
Day 3 | 5 | +5 | (+5 / 5 ) * 100 = +100% |
Day 4 | 7 | +2 | (+2 / 5) * 100 = 40% |
Day 5 | 2 | -5 | (-5 / 7 ) * 100 = -71.42% |
The average percentage change in daily ticket creation for a cluster is then calculated in this manner:
Total percentage / (number of days - 1)
In this case, -100 + 100 + 40 – 71.42 / 5-1 = -31.42 / 4 = -7.85%.