One-hot knowledge check

Below is a frequency table from a categorical pandas Series for a person's favorite soda.

Coke      1221
Pepsi     1117
NaN         21

If one-hot encoding is completed only on this column, and this is the only column in the dataset, how many columns will be in the final dataset?

Example code:

soda_onehot = pd.get_dummies(soda[["favorite_soda"]], columns=["favorite_soda"])

One: if the new column has a 0 for Coke, this means that their favorite soda was Pepsi.
Two: one for Coke and one for Pepsi. We can tell if a person did not respond if both columns are 0.
Three: one for Coke, one for Pepsi, and one for missing values.
Four: one for Coke, one for Pepsi, one for missing values, and one for the original column.