✨ np.array interface implementation

[jonasvdd]

Changes applied:

• removed the pd.Series dependency for the Aggregators and worked towards a numpy array Aggregation interface;

As a result: The aggregation_interface.py class structure is now:

• AbstractAggregator superclass; implemented by
  • DataPointSelector : Selects indices (will be the superclass for tsdownsample aggregators)
    • contains the arg_downsample method
  • DataAggregator: Aggregates the x and y
    • contains the aggregate method.

• EfficientLTTB -> MinaxLTTB
• Moved some functionality from the FigureResamplerInterface its _check_update_trace_data to a new (static) class PlotlyAggregatorParser
• Closes Could you pls relax pandas requirement pandas = "^1.3.5"? pandas==2.0.0 was released! #193

[jvdd]

Ready for review @jonasvdd?

[jonasvdd]

Now it is! @jvdd 🎉

[codecov-commenter]

Codecov Report

Merging #154 (497b2a2) into main (b1225bd) will increase coverage by 0.43%.
The diff coverage is 98.15%.

📣 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             @@
##             main     #154      +/-   ##
==========================================
+ Coverage   97.00%   97.44%   +0.43%     
==========================================
  Files          12       13       +1     
  Lines         902      978      +76     
==========================================
+ Hits          875      953      +78     
+ Misses         27       25       -2     

Impacted Files	Coverage Δ
...tly_resampler/aggregation/gap_handler_interface.py	95.00% <95.00%> (ø)
...tly_resampler/aggregation/aggregation_interface.py	95.00% <96.00%> (-3.47%)	⬇️
..._resampler/aggregation/plotly_aggregator_parser.py	96.62% <96.62%> (ø)
plotly_resampler/__init__.py	84.61% <100.00%> (ø)
plotly_resampler/aggregation/__init__.py	100.00% <100.00%> (ø)
plotly_resampler/aggregation/aggregators.py	100.00% <100.00%> (+8.33%)	⬆️
plotly_resampler/aggregation/gap_handlers.py	100.00% <100.00%> (ø)
...tly_resampler/figure_resampler/figure_resampler.py	89.16% <100.00%> (ø)
...ler/figure_resampler/figure_resampler_interface.py	100.00% <100.00%> (+0.26%)	⬆️
...sampler/figure_resampler/figurewidget_resampler.py	99.03% <100.00%> (ø)
... and 1 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jvdd]

Did a final review, will push a commit with some minor updates.
If we address the comments from this review, we can merge this one 🚀

---

Some general comments:

• Add tests for passing None when using range-index
• We should profile the code for unnecessary overhead
• Still not 100% sure whether we should have pandas as a core dependency 🤔

[jvdd]

This is also present in tsdownsample? 🤔

[jonasvdd]

will look into this!

[jonasvdd]

I woudl leave this as is for now; because:

1. tsdownsample it's interface has a hard lock-in on returning the selected indices; which is only true for the DataPointSelector classes.
2. And since, we also support data aggregation e.g. mean / ...; serving a dtype regex to that function might be beneifical.

Main thing still left to do:

• look if there is any unnecessary overhead w.r.t. the dtype regex (e.g. we define them for both the plotly-resampler and tsdownsample interface.) -> maybe we can just reuse tsdownsample its interface for the datapoint-selector` class?

[jvdd]

Ticked off all TODOs.
I'm fine with atm not having versioned docs - is mostly relevant for LTTB.c code (which is now removed and replaced by tsdownsample)

@jonasvdd, could you please review this pull request one final time? If everything looks good, we can merge it into the main branch. After that, we can consider publishing a release candidate. 🚀