Open val-ismaili opened 8 months ago
The issue of significance also applies to other Key Performance Indicators (KPIs) that consist of aggregated data. Frequently, when comparing these aggregated values, such as averages, the outcomes can seem unexpected, even counter intuitive. This situation may arise if outliers are affecting the combined data in ways that were not anticipated.
Current structure of
intermediate-pt-wait-time.csv
isCan we add a column that logs the count of wait events at each stop in that time period? This is useful context from a transport perspective providing context on significance of waiting times.
Also allows for sense-checking ( / "show your working") of main
kpi-pt-wait-time.csv
. Currently taking the mean of mean column in the intermediate output will not provide the same number as we would need to weight by count of 'waits'.