Function generate_dict_table

python/benchmarks/parquet.py:69–75

(num_cols, string_size, nunique, length,
                        random_order=True)

def generate_dict_table(num_cols, string_size, nunique, length,
                        random_order=True):
    data = generate_dict_strings(string_size, nunique, length,
                                 random_order=random_order)
    return pa.table([
        data for i in range(num_cols)
    ], names=[f'f{i}' for i in range(num_cols)])


class ParquetWriteDictionaries(object):