Weight aes generates ValueError: assignment destination is read-only

[krivard]

Hey, I was hoping to use plotnine to make a weighted histogram, but it looks like the weight aesthetic isn't working right. This may be related to pandas' shift to returning read-only numpy arrays -- you may be missing an explicit copy.

Given:

import plotnine as p9
import pandas as pd

df = pd.DataFrame({
    "x": list(range(10)),
    "w": list(range(10)),
})

(
    p9.ggplot(df, p9.aes(x="x", weight="w"))
    +p9.geom_histogram(bins=10)
)

Expected:

A histogram with ten bins, starting low and getting higher from left to right

Actual:

[...]
File [.../python3.12/site-packages/plotnine/stats/binning.py:169](.../python3.12/site-packages/plotnine/stats/binning.py#line=168), in assign_bins(x, breaks, weight, pad, closed)
    167 else:
    168     weight = np.asarray(weight)
--> 169     weight[np.isnan(weight)] = 0
    171 bin_idx = pd.cut(
    172     x,
    173     bins=breaks,  # type: ignore
   (...)    176     include_lowest=True,
    177 )
    178 bin_widths = np.diff(breaks)

ValueError: assignment destination is read-only