Adding uniqueness with Sorted Sets and HyperLogLog

This section presents two different Time Series implementations that support unique insertions (for example, unique visitors or concurrent video plays), which are very similar to the previous solutions.

The first implementation uses Sorted Sets, and it is based on the previous Hash implementation. The second implementation uses HyperLogLog, and it is based on the previous String implementation. Since these new implementations are very similar to previous ones, only the lines highlighted in bold are explained.

Each solution has pros and cons:

• The Sorted Set solution works well and is 100% accurate
• The HyperLogLog solution uses less memory than the Sorted Set solution, but it is only 99.19% accurate

The proper solution should be chosen based on how much data needs to be stored and how accurate it needs to be.

Create a file called timeseries-sorted-set.js, copy the content of timeseries-hash.js, and change the following:

function TimeSeries(client...