Though Facet's primary application is analyzing time series datasets, a dedicated time series database is not required to use Facet. Through rigorous benchmarking, we've concluded that Facet works as an effective analytics tool for all modern cloud data warehouses at most data volumes. This means that you can use Facet as a powerful, analytics layer atop your existing cloud data warehouse, without the hassle of maintaining a complex database.
The only requirement a dataset must have to be used in Facet is that it contain at least one Date or Datetime column.
If your table meets this criterion, you can easily begin to explore the data in a Facet workspace.
We have a few guidelines around what the ideal Facet dataset looks like, which you can follow to make sure you get the most value from your workspace.
Facet is able to effectively visualize a dataset of any size, no matter how big or small. There are use cases for Facet which contain thousands of events per day, and other use cases which have tens of billions of events per day. What's most important is having enough data where the dataset is continuous across its time frame. You will be able to extract more value from the tool when all of the time series chart lines are continuous.
Inside the Facet workspace, each dataset has two key building blocks: dimensions and metrics.
- A dimension is a data attribute that describes your data: Country, Device OS, or Account Manager
- A metric is a quantitative measurement derived from your data: Impressions, Clicks, or Revenue
An ideal dataset for Facet has many dimensions and many metrics. We find that datasets with at least 10 dimensions and 5 metrics are best, but this really comes down to the use case being solved.
Updated 7 months ago