Cache ecobenefits for a single tree

[sireliah]

Overview

The /tree-benefits endpoint returns NO3, O3, PM2.5, etc. values for a single tree (based on the trunk diameter in the POST request). However we are making heavy model calculations every time we get request. We should cache the response in Redis.

Endpoint

https://github.com/kodujdlapolski/ecobenefits/blob/master/eco/server.py#L48
https://github.com/kodujdlapolski/ecobenefits#tree-benefits-endpoint

What to do

For the each request for given trunk diameter check if the result is already in Redis. If not, then use predict_tree_benefits function and store results in Redis.

Redis connection can be obtained here: https://github.com/kodujdlapolski/ecobenefits/blob/master/eco/server.py#L22. Asynchronous connection is preferred, but if there are problems, ordinary synchronous can be also used: https://github.com/kodujdlapolski/ecobenefits/blob/master/eco/utils.py#L12

It would be also nice to write some API tests for the /tree-benefits endpoint. Maybe this library can help: https://github.com/Martiusweb/asynctest. If testing of the whole endpoint is too complicated (too much to mock, async problems), then please write unit tests.