Import aggregated timeseries code

[mhidas]

Import code from data-services and incorporate into the package.

[codecov[bot]]

Codecov Report

Merging #52 into master will decrease coverage by 10.55%. The diff coverage is 76.13%.

@@             Coverage Diff             @@
##           master      #52       +/-   ##
===========================================
- Coverage     100%   89.44%   -10.56%     
===========================================
  Files           3        4        +1     
  Lines         222      398      +176     
  Branches       36       63       +27     
===========================================
+ Hits          222      356      +134     
- Misses          0       26       +26     
- Partials        0       16       +16

Impacted Files	Coverage Δ
...tools/timeseries_products/aggregated_timeseries.py	76.13% <76.13%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 1dc0146...c7299f2. Read the comment docs.

[mhidas]

@diodon FYI. Not complete yet.

[mhidas]

@diodon This is ready to review/merge. I've added a quick unittest for the product code. It's far from complete, but good enough for now.

@lwgordonimos FYI. I'd like to merge this today. Feel free to take a quick look either before or after that if you have a minute.

[mhidas]

BTW there's more refactoring to come, and adding documentation, in other PRs.

[ghost]

FYI, this doesn't work on Python 2 without pinning some libraries, since they have dropped Python 2 support completely at these versions:

    'numpy>=1.13.0,<1.17.0',
    'pandas<0.25.0',
    'xarray<0.12.0'

Hopefully a very short term pinning!

[mhidas]

:+1: Thanks @lwgordonimos I'll add those in my next PR