Overview
The Daily Event Count calculator analyzes event frequency patterns by grouping events by their date, providing daily event count distribution and percentage analysis. This calculator counts how many events occurred on each calendar day across your entire event log, making it useful for identifying data quality issues, workload patterns, and system activity trends.
Unlike case-level calculators, this operates at the event level, meaning it counts individual activities or transactions rather than complete process instances.
Common Uses
- Identify data extraction gaps or missing days in your event log
- Detect unusual spikes or drops in system activity that may indicate data quality issues
- Analyze workload distribution across calendar days
- Understand weekend versus weekday activity patterns
- Identify seasonal trends and patterns in process execution
- Validate data completeness and consistency over time
Settings
This calculator has no configurable settings. It automatically analyzes all events in your filtered dataset by grouping them by the date component of their timestamp.
Standard Fields:
- Title: Optional custom title for the calculator output
- Description: Optional description for documentation purposes
Examples
Example 1: Detecting Data Extraction Issues
Scenario: You're validating a new data extraction from your ERP system and want to ensure that events are being captured every business day without gaps.
Settings:
- Title: "Daily Event Distribution"
- Description: "Verify completeness of data extraction"
Output:
The calculator displays a table with three columns:
- Date: Each calendar day found in the event log
- Count: The number of events that occurred on that day
- Percent: The percentage of total events that occurred on that day (as a decimal)
Example output:
Date Count Percent
2024-01-15 1,247 0.0523
2024-01-16 1,189 0.0499
2024-01-17 42 0.0018
2024-01-18 1,312 0.0551
Insights: In this example, January 17th shows only 42 events compared to the typical 1,200+ events on surrounding days. This dramatic drop (less than 2% of normal volume) indicates a potential data extraction problem or system outage that should be investigated. Look for missing days (gaps in the date sequence) or days with unusually low counts that might indicate incomplete data.
Example 2: Weekend vs Weekday Analysis
Scenario: You want to understand whether your invoice processing activities occur seven days a week or only on business days.
Settings:
- Title: "Invoice Processing Activity Calendar"
- Description: "Identify processing patterns across the week"
Output:
The daily distribution shows event counts for each day. When visualized as a chart, you can identify:
- Days with zero events (likely weekends and holidays)
- Consistent weekday patterns
- Monday spikes (common in many business processes)
- End-of-month peaks
Insights: If you see near-zero event counts on Saturdays and Sundays, your invoice processing is primarily a weekday activity. If you see activity seven days a week, you may have automated processing or international operations. Significant Monday spikes often indicate backlog processing from the weekend.
Example 3: Identifying System Upgrade Impact
Scenario: Your IT department performed a system upgrade on March 15th, and you want to verify whether it impacted transaction processing volume.
Settings:
- Title: "March System Activity Analysis"
- Description: "Before and after upgrade comparison"
Output:
The calculator shows event counts for each day in March. You can compare the average daily count before March 15th with the average daily count after March 15th.
Example pattern:
Date Count Percent
2024-03-12 2,450 0.0334
2024-03-13 2,387 0.0325
2024-03-14 2,512 0.0342
2024-03-15 876 0.0119 <- Upgrade day
2024-03-16 2,398 0.0327
2024-03-17 2,441 0.0332
Insights: The significantly lower event count on March 15th (876 vs typical 2,400+) shows reduced system activity during the upgrade window. The return to normal volumes on March 16th indicates the system recovered successfully. If low counts continued for several days, it would suggest post-upgrade issues requiring investigation.
Output
The calculator produces a data table with the following columns:
Date (DateTime): The calendar date (without time-of-day component) for each group. Results are ordered chronologically from earliest to latest date.
Count (Number): The total number of events that occurred on that specific date. This counts all activities/events in your event log for that day.
Percent (Decimal): The percentage of total events represented by that date, shown as a decimal value (for example, 0.15 represents 15% of all events).
The output can be visualized as:
- Line charts: Ideal for showing daily trends over time
- Bar charts: Effective for comparing activity levels across specific date ranges
- Calendar heat maps: Visual representation of activity intensity by date
- Time series analysis: For identifying trends and patterns over longer periods
Note: Events with missing or invalid timestamps are excluded from the analysis. Only events with valid date/time information are counted.
This documentation is part of the mindzie Studio process mining platform.