Financial timeseries analysis brainstorming

There are a lot of ways datashader could be used for analyzing timeseries data, and it would be useful to run a project where we work closely with a financial (e.g.) analyst to prototype and then implement different approaches.

For instance, one may want to select a specific timeseries, then show a datashaded plot of all the timeseries that covary with the first one (according to some similarity metric, and some threshold value). Or we may need to brainstorm around techniques for cyclical data (issue #105), in the specific context of finanical data. This issue is here just to keep track of such ideas until we get around to being able to work on them.