Some interesting real world anomalies from various business metrics.
Note: May need to right click and open image in new tab to see them properly.
bump_and_spike_example: Nice example with both a spread out bump and a spike.etl_delay_example: Example where a regular ETL process changes unexpectedly. The metric in this case was minutes since table was last updated, the anomaly was the regular update schedule being missed.increasing_table_row_count_then_sudden_drop: This was metric of row counts on a table expected to keep increasing over time. The anomaly was a sudden drop in row count. This was related to some backend cleaning scripts that were run but turned out very important to catch/understand when interpreting the data.saw_tooth_example: Example of some increased variability leading to a "saw tooth" pattern.sharp_drop_example: Example with a sharp drop leading to an anomaly.
