use spark-shell to group data and find mean on distinct dates from json files

Question

I have a directory with 10 days of separate data Folders. Each Date folder has a JSON files as following.

[{"value": 5,"count" : 16,"curr