Adding SimpleRidgeRegressor

[ayush-1506]

This is the first part in a series of fixes for #125.
Changes include:

1. SimpleRidgeRegressor implementation.
2. Modification of tests and docs.

@ablaom Since this is my first PR, I'd love a review.

[codecov-io]

Codecov Report

❗ No coverage uploaded for pull request base (master@3d3c4f9). Click here to learn what that means.
The diff coverage is 66.66%.

@@            Coverage Diff            @@
##             master     #129   +/-   ##
=========================================
  Coverage          ?   67.57%           
=========================================
  Files             ?       19           
  Lines             ?     1252           
  Branches          ?        0           
=========================================
  Hits              ?      846           
  Misses            ?      406           
  Partials          ?        0

Impacted Files	Coverage Δ
src/MLJ.jl	100% <ø> (ø)
src/builtins/ridge.jl	66.66% <66.66%> (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 3d3c4f9...9e18a91. Read the comment docs.

[giordano]

@ablaom if you don't want coveralls to create a comment in each PR, you can disable this feature in the settings: https://coveralls.io/github/alan-turing-institute/MLJ.jl/settings

Edit: just realised that the comment above is by codecov, which is more annoying to mute

[ablaom]

Excellent, thanks. Some minor points:

• in the test you should be giving fit a table not a matrix. The input X is always a table (unless input_is_multivariate(ModelType) = false). So just put ip = MLJ.table(ip') for the test.

• make the name of the test file the same as the file being tested: test/SimpleRidgeRegressor -> test/ridge.jl

[ablaom]

Thanks for that!

There were some changes to LocalMultivariateStats.jl in tests that we didn't want, which I fixed.