This is a self-guided tutorial to FS2, using streaming data analytics for the majority of examples. It's a combination of text and coding exercises.
This is the rough plan for contents:
- FS2
Stream
as akin toList
- The
Stream
API works likeList
, except:- contents are (potentially) arranged in time not space
- separation between description and action (as in
IO
) so need to "compile
" theStream
- Basic examples: map, filter, etc.
- Calculating averages
- Naive approach
- Incremental approach (e.g. https://math.stackexchange.com/questions/106700/incremental-averaging)
- EWMA
- The
- Effects in FS2
- Fan-in and fan-out
- Pull based model
- Tools for each
- Streaming algorithms
- Single pass
- Streaming average
- Reservoir sampling
- Hyperloglog
- Sliding and tumbling windows