MNT: Modularize Rocket Draw

[MateusStano]

Pull request type

• Code maintenance (refactoring, formatting, tests)

Checklist

• Lint (black rocketpy/ tests/) has passed locally
• All tests (pytest tests -m slow --runslow) have passed locally
• CHANGELOG.md has been updated (if relevant)

Current behavior

The method of _RocketPlots relating to the rocket drawing was too big and hard to change/improve/maintain

New behavior

The .draw() method has been separated into several different methods. No code relating to the plot has been changed.

Breaking change

• No

Additional information

I was trying to add sensors to the drawing and it was really hard to work with the drawing, so I refactored it.

A quick note:

• There are a few lines of repeated code that could still be turned into functions. These are usually some matplotlib one-liners. I did not create a function for these because it affected readability negatively.

[codecov]

Codecov Report

Attention: Patch coverage is 94.28571% with 4 lines in your changes are missing coverage. Please review.

Project coverage is 72.94%. Comparing base (bb0f46a) to head (822a89e).
Report is 7 commits behind head on develop.

❗ Current head 822a89e differs from pull request most recent head 2cbaa77. Consider uploading reports for the commit 2cbaa77 to get more accurate results

Files	Patch %	Lines
rocketpy/plots/rocket_plots.py	94.28%	4 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #580      +/-   ##
===========================================
- Coverage    72.99%   72.94%   -0.06%     
===========================================
  Files           59       59              
  Lines         9543     9612      +69     
===========================================
+ Hits          6966     7011      +45     
- Misses        2577     2601      +24     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Gui-FernandesBR]

Overall, very good PR. I understand you refactored the code without changing the behavior. You probably ran all the example notebooks to check the plots are being rendered correctly.

Some lines in the _draw_nozzle_tube function are missing coverage.
Could you please take a look at that? Just to ensure the code never brakes.

[Gui-FernandesBR]

This is strange to me. You add the surface to a list, which was passed as an argument to the function. If you don't return the list with the added surfaces, is this code line really necessary? I'm afraid that everytime you are starting with an empty list.

        # Add the nosecone to the list of drawn surfaces
        drawn_surfaces.append(
            (surface, x_nosecone[-1], surface.rocket_radius, x_nosecone[-1])
        )

[Gui-FernandesBR]

This behavior is also present in other parts of the code.

[MateusStano]

The drawn_surfaces list is created and then appended to in the various draw methods I created. That same list is used later to draw the tubes later.

Remember that a list in Python is just a pointer. Changing the name of the list variable does not change the address that it points to.

[Gui-FernandesBR]

Also, @MateusStano can you please test the implementation in a multi-stage rocket? I think the NDRT example notebook may help you with that.

[MateusStano]

Also, @MateusStano can you please test the implementation in a multi-stage rocket? I think the NDRT example notebook may help you with that.

The drawing ends up identical to the way it was before. Did you find any problems there?

[Gui-FernandesBR]

Also, @MateusStano can you please test the implementation in a multi-stage rocket? I think the NDRT example notebook may help you with that.

The drawing ends up identical to the way it was before. Did you find any problems there?

Nope. Sorry, I just did not have enough time to checkout your branch and run the examples.

If you have tested all the examples, I trust you. Go ahead and merge it.

[Gui-FernandesBR]

Good work.