This PR illustrates the idea mentioned in #244 (comment) which is motivated by the fixest::i() syntax that gives better control over interactions of factor variables.

The only change is the addition of the attribute drop in TreatmentContrasts which allows the user to specify if the reference level in the contrast encoding should be dropped.

For example,

rng = np.random.default_rng(91)
df = pd.DataFrame(
    {
        "factor1": rng.choice(["a", "b", "c"], 10),
        "factor2": rng.choice([1, 2, 3], 10),
        "y": rng.normal(0, 1, 10),
    },
)

model_matrix(
    "C(factor1, contr.treatment('a', drop=True))",
    data=df,
    ensure_full_rank=False,
)

(cc @s3alfisc)