Business Machine Learning/

...

/

Solution: Retrain the Model Without Pclass Dummies

Solution: Retrain the Model Without Pclass Dummies

View the solution to the “Retrain the Model Without Pclass Dummies” exercise.

We'll cover the following...

Let's practice what we have learned so far.

Task

The task here is to retrain the model with the original Pclass column without dummies and obtain the model results.

Solution

The provided workspace contains the code solution for the task mentioned above.

Python 3.8
# loading the train data
train = pd.read_csv("train_titanic_Xy.csv")
# pre-processing
def impute_age(age_pclass):
    Age = age_pclass[0]
    Pclass = age_pclass[1]
    if pd.isnull(Age):
        if Pclass == 1:
            return 38
        elif Pclass == 2:
            return 30
        else:
            return 25
    else:
        return Age
train["Age"] = train[["Age", "Pclass"]].apply(impute_age, axis=1)
train.drop("Cabin", axis=1, inplace=True)
train.dropna(inplace=True)
# creating dummies
sex_emb = pd.get_dummies(train[["Sex", "Embarked"]], drop_first=True)
train = pd.concat([train, sex_emb], axis=1)
# get rid of extraneous features
train.drop(["Sex", "Embarked", "Name", "Ticket", "PassengerId"], inplace=True, axis=1)
# training data
X_train = train.drop("Survived", axis=1)
y_train = train["Survived"]
# scaling the data
scaler = MinMaxScaler()
scaler.fit(X_train)
X_train_s = scaler.transform(X_train)
# model ==> training
logR = LogisticRegression(penalty="l2", max_iter=10000, C=0.5, n_jobs=-1)
logR.fit(X_train_s, y_train)
# loading the test data
X_test = pd.read_csv("test_titanic_X.csv")
y_test = pd.read_csv("test_titanic_y.csv")
test = pd.merge(X_test, y_test, on="PassengerId", how="left")
# pre-processing
test["Age"] = test[["Age", "Pclass"]].apply(impute_age, axis=1)
test.drop("Cabin", axis=1, inplace=True)
test.dropna(inplace=True)
# repeat
sex_emb_test = pd.get_dummies(test[["Sex", "Embarked"]], drop_first=True)
test = pd.concat([test, sex_emb_test], axis=1)
test.drop(["Sex", "Embarked", "Name", "Ticket", "PassengerId"], axis=1, inplace=True)
# testing data
X_test = test.drop("Survived", axis=1)
y_test = test["Survived"]
# scaling the data
X_test_s = scaler.transform(X_test)
X_test_s = pd.DataFrame(X_test_s, columns=X_test.columns)
# predictions
pred_train = logR.predict(X_train_s)
pred_test = logR.predict(X_test_s)
# display
print("Training score (mean accuracy): ", logR.score(X_train_s, y_train))
print("Test Score (mean accuracy): ", logR.score(X_test_s, y_test))
...