fix(geom_path): allow layers without data

[raashish1601]

Plotting an empty data frame with a path based layer raises a KeyError:

data = pd.DataFrame({"x": pd.Series([], dtype=float)})
ggplot(data, aes("x")) + stat_ecdf()
# KeyError: "['y'] not in index"

The same happens with geom_freqpoly and geom_qq_line. When a stat gets no rows it returns no computed columns, so the layer data has no y. geom_path.handle_na then selects ["x", "y", "size", "color", "linetype"] strictly. Other geoms (e.g. geom_histogram, geom_density) already draw nothing in this case, and ggplot2's GeomPath$handle_na only looks at the columns that are present (data[names(data) %in% c(...)]).

This makes handle_na use only the columns that exist.

Testing: added test_empty_layer_data, which fails on main and passes with the fix. tests/test_geom_path_line_step.py, tests/test_stat_ecdf.py, tests/test_geom_freqpoly.py and tests/test_geom_qq_qq_line.py pass, ruff is clean. Added a changelog entry under Bug Fixes.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 88.87%. Comparing base (ac9cfce) to head (630b619).
⚠️ Report is 4 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1136      +/-   ##
==========================================
- Coverage   88.87%   88.87%   -0.01%     
==========================================
  Files         229      229              
  Lines       17292    17291       -1     
  Branches     2203     2203              
==========================================
- Hits        15368    15367       -1     
  Misses       1295     1295              
  Partials      629      629              

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.