Fix numeric hf_y input as dtype object

[jonasvdd]

Should fix #89

TODO:

• add a test

[codecov-commenter]

Codecov Report

Merging #90 (3d8e3b5) into main (cc951c1) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main      #90   +/-   ##
=======================================
  Coverage   98.12%   98.12%           
=======================================
  Files          10       10           
  Lines         745      746    +1     
=======================================
+ Hits          731      732    +1     
  Misses         14       14           

Impacted Files	Coverage Δ
...tly_resampler/aggregation/aggregation_interface.py	98.46% <ø> (-0.05%)	⬇️
...ler/figure_resampler/figure_resampler_interface.py	99.70% <100.00%> (+<0.01%)	⬆️

---

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 cc951c1...3d8e3b5. Read the comment docs.

[jvdd]

Hmmm not really keen on this implementation. When you have data type object that contains bools it gets now converted to float as well (which is not in line with the plotly behavior).

fig = go.Figure()
fig.add_trace(
    go.Scattergl(
        y=np.array([True, False]*20).astype(object),
        mode="markers"
    )
)
# This shows true & false on the y-axis

fig = FigureWidgetResampler()
fig.add_trace(
    go.Scattergl(
        mode="markers"
    ),
    hf_y = np.array([True, False]*20).astype(object)
)
# This shows 0 & 1 on the y-axis

[jonasvdd]

Maybe first try parsing to bool?

[jvdd]

LGTM! Added some extra tests for binary data (to check if it does not get transformed to bool)