If you have ever tried to use native python objects in Pandas DataFrames, you may have run into an issue similar to this:
df = pd.DataFrame({"A": [1, 2, 3, 4],
"B": ["a", "b", "c", "d"],
"C": [[1, 2, 3], [1, 2], [1, 2, 3], 4],
"D": [{1: 1, 2: 2}, {1: 1, 3: 3}, {1: 1, 4: 4}, {1: 1, 2: 2}],
"E": [[{1: {2: 2}}, {2: {3: 3}}], [{1: {2: 2}}, {2: {3: 3}}],
[{1: {2: 2}}, {2: {3: 3}}], [{1: {2: 2}}, {2: {3: 3}}]]
})
df['C'].unique()TypeError: unhashable type: 'list'
This is caused by unhashable values in the DataFrame cells.
This small library helps to resolve that making this possible:
from hashable_df import hashable_df
hashable_df(df)['E'].unique()returning
array([[{1: {2: 2}}, {2: {3: 3}}]], dtype=object)
