understanding how onehotencoder works - why do i get mutliple ones in ohe column?

Question

I am using sklearn pipelines to perform one-hot encoding:

preprocess = make_column_transformer(
    (MinMaxScaler(),numeric_cols),
    (OneHotEncoder(),['country'])
    )

param_grid =    { 
                  'xgbclassifier__learning_rate': [0.01,0.005,0.001],
                 
                  }

model = make_pipeline(preprocess,XGBClassifier())

# Initialize Grid Search Modelg
model = GridSearchCV(model,param_grid = param_grid,scoring = 'roc_auc',
                                 verbose= 1,iid= True,
                                     refit = True,cv  = 3)
model.fit(X_train,y_train)

To see then how the countries are one hot encoded I get the following ( I know there are two)

pd.DataFrame(preprocess.fit_transform(X_test))

The result of this is:

A few questions:

• now correct me if wrong but in one hot encoding I thought it was a series of all 0's and just ONE number 1. why do I get several ones in one column
• when I do model.predict(x_test) it applies the trasnformations as defined in the piepline fom training?
• how do I retrieve the feature names when I call fit_transform?

Answer 1

To help you better understand (1), i.e. how OHE works.

Suppose you have 1 column with categorical data:

df = pd.DataFrame({"categorical": ["a","b","a"]})
print(df)
  categorical
0           a
1           b
2           a

Then you'll get one 1 per row (this will always be true for one column categorical data), but not necessarily on a per column basis:

from sklearn.preprocessing import OneHotEncoder
ohe = OneHotEncoder()
ohe.fit(df)
ohe_out = ohe.transform(df).todense()
# ohe_df = pd.DataFrame(ohe_out, columns=ohe.get_feature_names(df.columns))
ohe_df = pd.DataFrame(ohe_out, columns=ohe.get_feature_names(["categorical"]))
print(ohe_df)
   categorical_a  categorical_b
0            1.0            0.0
1            0.0            1.0
2            1.0            0.0

Should you add more data columns, e.g. a numerical column, this will hold true on a per column basis, but not for the whole row anymore:

df = pd.DataFrame({"categorical":["a","b","a"],"nums":[0,1,0]})
print(df)
  categorical  nums
0           a     0
1           b     1
2           a     0

---

ohe.fit(df)
ohe_out = ohe.transform(df).todense()
# ohe_df = pd.DataFrame(ohe_out, columns=ohe.get_feature_names(df.columns))
ohe_df = pd.DataFrame(ohe_out, columns=ohe.get_feature_names(["categorical","nums"]))
print(ohe_df)
   categorical_a  categorical_b  nums_0  nums_1
0            1.0            0.0     1.0     0.0
1            0.0            1.0     0.0     1.0
2            1.0            0.0     1.0     0.0