feat: Add "OTHER" category to the encoders when min_frequency is supplied

[vspinu]

Currently (as per v0.1.3) categorical encoders don't handle the other categories when min_frequency is supplied. For example:

    import pandas as pd
    import ibis
    import ibis_ml as ml

    con = ibis.duckdb.connect()

    df = pd.DataFrame({
        'cat1': ['AA', 'BBB', 'AA', 'BBB', 'CCC'],
        'cat2': ['X', 'Y', 'Y', 'X', 'Z'],
        'value': [10, 20, 30, 40, 50]
    })

    tbl = con.create_table("tmp", df, overwrite=True)
 
    tr_ohe = ml.Recipe(
        ml.OneHotEncode(ml.string(), min_frequency=2),
    ).fit(tbl.drop("value"), tbl.value)
    
    tr_ohe.to_ibis(tbl).to_pandas()
     #    value  cat1_AA  cat1_BBB  cat2_X  cat2_Y
    # 0     10        1         0       1       0
    # 1     20        0         1       0       1
    # 2     30        1         0       0       1
    # 3     40        0         1       1       0
    # 4     50        0         0       0       0


    tr_oe = ml.Recipe(
        ml.OrdinalEncode(ml.string(), min_frequency=2),
        ml.FillNA(ml.integer(), "OTHER")       # <-- this does not work!!! 
    ).fit(tbl.drop("value"), tbl.value)
    
    tr_oe.to_ibis(tbl).to_pandas()
   #    value  cat1  cat2
    # 0     10   0.0   0.0
    # 1     20   1.0   1.0
    # 2     30   0.0   1.0
    # 3     40   1.0   0.0
    # 4     50   NaN   NaN

(Note that there is an additional issue that filling NaN does not work on int vectors)

Would it be possible to extend the encoders to be able to specify the "other" value? More concretely:

• For OHE add the ability to generate the column which is 1 for all the swallowed categories.
• For OHE and OE add "others_value" argument to indicate which value should the swallowed categories assume