Index overlap on multiindex with auto_fit_columns

[maltomato]

Summary

When I create a grid with a multiindex and apply auto_fit_columns the columns overlap. If the error occurs depends on the values. Some index value combinations lead to the issue and others do not.

To reproduce

This version creates the error:

import ipydatagrid
import numpy as np
import pandas as pd

arrays = [
    [1, 1, 1],
    [2, 2.5, 2.5],
    [999, 999, 999],
    [1000, 1000, 1000],
]
index = pd.MultiIndex.from_arrays(arrays, names=('first', 'second', 'third', 'fourth'))
columns = ['A', 'B', 'C', 'D']
df = pd.DataFrame(np.random.randn(3, 4), index=index, columns=columns)

grid = ipydatagrid.DataGrid(df, auto_fit_columns=True)
grid

With a slight change in the index values (changing the third 2.5 on index level 2) and it looks normal again.

import ipydatagrid
import numpy as np
import pandas as pd

arrays = [
    [1, 1, 1],
    [2, 2.5, 2],
    [999, 999, 999],
    [1000, 1000, 1000],
]
index = pd.MultiIndex.from_arrays(arrays, names=('first', 'second', 'third', 'fourth'))
columns = ['A', 'B', 'C', 'D']
df = pd.DataFrame(np.random.randn(3, 4), index=index, columns=columns)

grid = ipydatagrid.DataGrid(df, auto_fit_columns=True)
grid

System info

• ipydatagrid 1.3.2
• jupyterlab 4.2.5
• python 3.12.5
• ipython 8.27.0