Guides
Guides

Sample count in the query response

The number of observations recorded to produce each record (line) in the data file, as a non-negative integer. The meaning of this field depends on the kind of sample data requested, specifically probe data or path data. For more information, see the sections on Path Data and on Per Vehicle or Per Probe. If you requested Probe data, then this is the number of probe samples collected from all observed vehicles on the road segment in the epoch. Different vehicles produce samples at different frequencies, depending on the sampling technology used. Also, a vehicle may collect more than one sample in an epoch as it transits the road segment. If you need to aggregate several Traffic Analytics records, you can use this field to weight the aggregation by the number of observations.

If you have selected probe data “per vehicle” in your query, then this field represents the number of unique sampled vehicles which used the segment in the epoch, rather than the number of probe samples. However, there is a limitation in this use case. The “per vehicle” counts are valid only when you select 5-minute epochs. If you select longer epochs (15-minute or 60-minute) then this field is the sum of the unique vehicles observed in each 5-minute epoch, which is usually greater than the number of unique vehicles across the larger time span.

If you request path data, then this field is the number of vehicle paths which crossed the road segment in the epoch. See the section on path versus probe in the previous section for more information on the differences and use cases. With path counts, as with probe per vehicle counts, only 5-minute epochs correctly give a count of unique vehicles.

And if you request both path and probe data, then this field gives the sum of the two values previously detailed. The sample count field doesn't represent the total number of vehicles which crossed the road segment (“vehicle volumes”), in any configuration. It only represents the vehicles which transmitted probe samples to HERE. You can draw some limited inferences about relative vehicle volumes from a Traffic Analytics dataset, but there are several factors which limit the precision of such inferences. For example, system outages in the sampling infrastructure count lead to temporarily low sample counts. Also, your license agreement for Traffic Analytics may limit the use of the count field to produce vehicle volume data.