Improve error message for ColumnFormula constraint when constraint column used in formula

[fealho]

Using the ColumnFormula constraint where the constraint column is used in the constraint formula crashes with a KeyError. While this is expected, the error message could be improved to provide a more meaningful explanation of what's causing the issue.

Below is some code to produce this error:

import pandas as pd
from sdv.constraints import *
from sdv.tabular import *

ones = [1 for i in range(10)]
data = pd.DataFrame({
    'a': ones,
    'b': ones,
    'c': ones
})

def formula(data):
    return data['a']

constraint = ColumnFormula(
    column='a',
    formula=formula
)

model = GaussianCopula(constraints=[constraint])
model.fit(data)
samples = model.sample(10)

[katxiao]

Instead of updating the comment, we will add an argument to optionally keep the original column instead of always dropping it.