Fix for get_matching_df for latest Pandas version

in stages/track_building/utils.py, get_matching_df as written does not work for latest Pandas version (2.2.2). This can be fixed by replacing the definitions of candidate_lengths, and particle_lengths with:

    candidate_lengths = (
        reconstruction_df.track_id.value_counts(sort=False)
        .reset_index()
        .rename(columns={"count": "n_reco_hits"})
    )

    particle_lengths = (
        reconstruction_df.drop_duplicates(subset=["hit_id"])
        .particle_id.value_counts(sort=False)
        .reset_index()
        .rename(columns={"count": "n_true_hits"})
    )