FloatFormatter does not round the data correctly for integer columns when using _set_fitted_parameters

[R-Palazzo]

Environment Details

• RDT version: 1.12.3

Error Description

We can't currently set the rounding_digits attribute of the FloatFormatter to 0 with _set_fitted_parameters. It seems to be because of the following line:

RDT/rdt/transformers/numerical.py

Line 239 in bb3262e

if rounding_digits:

Steps to reproduce

In the code below the reverse_transformed_data will not be rounded while rounding_digits=0

import pandas as pd
import numpy as np
from rdt.transformers import FloatFormatter, NullTransformer

data = pd.DataFrame({
    'col 1': 100 * np.random.random(10),
})
transformer = FloatFormatter()
transformer._set_fitted_parameters(
    column_name='col 1',
    null_transformer=NullTransformer(),
    rounding_digits=0,
    dtype='float',
)

reverse_transformed_data = transformer.reverse_transform(data)
reverse_transformed_data