Sample Size of PeMS Data sets

The destination SQL database contains a table-valued function [pems].[fn_sample_size] that returns a table of each PeMS data-set station by year with the number of observations and percentage of observations provided over the year. For example, a count station providing daily counts for the year 2014 with 300 records would show N = 300 and pct = 300/365.

This table can be queried at the station level to show what data exists and how much coverage the data-set has when creating yearly summary statistics. It can also be aggregated to show what data exists for time specific periods, within station or PeMS data-set.

It is important to always be aware of the sample size when creating summaries of the PeMS data as it is not guaranteed to be complete for any given time period.