Fix #225: Improve Display for naive_bayes

[m7142yosuke]

Fixes #225

Checklist

• My branch is up-to-date with development branch.
• Everything works and tested on latest stable Rust.
• Coverage and Linting have been applied

Current behaviour

Currently println!("{}", &gnb) prints:

GaussianNB:
inner: BaseNaiveBayes { distribution: GaussianNBDistribution { class_labels: [0, 1, 2], class_count: [50, 50, 50], class_priors: [0.3333333333333333, 0.3333333333333333, 0.3333333333333333], var: [[0.12176398698426993, 0.1422760001411465, 0.0295040034446723, 0.01126400058841702], [0.26110400788880384, 0.09650000076294418, 0.21639999752045114, 0.03832399889564586], [0.3962559386291673, 0.10192399745559833, 0.2984959836425922, 0.07392400431251556]], theta: [[5.006000003814697, 3.41800000667572, 1.4639999961853027, 0.24400000482797624], [5.935999975204468, 2.770000009536743, 4.259999980926514, 1.3259999918937684], [6.588000001907349, 2.9739999914169313, 5.551999988555909, 2.0259999775886537]] }, _phantom_tx: PhantomData, _phantom_ty: PhantomData, _phantom_x: PhantomData, _phantom_y: PhantomData }

New expected behaviour

GaussianNB:
 inner: BaseNaiveBayes: {
 distribution: GaussianNBDistribution: {
    class_labels: [1, 2],
    class_count: [3, 3],
    class_priors: [0.3, 0.7],
    var: [[0.666666666666667, 0.22222222222222232], [0.666666666666667, 0.22222222222222232]],
    theta: [[-2.0, -1.3333333333333333], [2.0, 1.3333333333333333]]
}}

Change logs

[m7142yosuke]

I'm just starting to learn rust, so I'm sorry if I'm wrong.
I implemented it differently from the format presented in #225 . Is it better to implement just like the example?

[Mec-iS]

thanks a lot!

could you try to do the same for other structures that implement Display? Could you find a way to avoid printing the _phantom fields?

[m7142yosuke]

@Mec-iS
I tried. Can you review it?

[Mec-iS]

Good job.

This could be merged but it will fix only the Gaussian, what about the other distributions? Please double check that all of them return well formatted output.

EDIT: Also please test your branch with smartcore-jupyter by using:
:dep smartcore = {git = "https://github.com/m7142yosuke/smartcore.git", , branch = "feature/improve-display-for-naive-bayes", features = ["datasets"]}
at the top of the notebooks. These changes are meant mostly for Jupyter pretty printing so it would be good to have a look how the branch works in the notebooks. For example see this.

[Mec-iS]

please see comments for desirable changes

[codecov-commenter]

Codecov Report

Merging #241 (19e5115) into development (d15ea43) will increase coverage by 0.07%.
The diff coverage is 81.81%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@               Coverage Diff               @@
##           development     #241      +/-   ##
===============================================
+ Coverage        44.50%   44.58%   +0.07%     
===============================================
  Files               85       85              
  Lines             7226     7234       +8     
===============================================
+ Hits              3216     3225       +9     
+ Misses            4010     4009       -1     

Impacted Files	Coverage Δ
src/naive_bayes/gaussian.rs	42.30% <81.81%> (+6.89%)	⬆️

... and 5 files with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.