KaggleTitanic_SimpleBlog.py

import pandas as pd
from sklearn.base import TransformerMixin
from xgboost import XGBClassifier
from sklearn.pipeline import Pipeline
from sklearn import metrics


class DataTransformer(TransformerMixin):  
    def cabin(self, val):
        if type(val) != str or val == "":
            return 0
        else:
            return 1
        
    def fit(self, X, y=None):
        assert isinstance(X, pd.DataFrame)
        data = X.copy()
        
        data = pd.get_dummies(data)
        # saving column names for maintaining consistent column names in validation/test data
        self.features = data.columns.values
        return self

    def transform(self, X, y=None):
        assert isinstance(X, pd.DataFrame)
        data = X.copy()
            
        data.loc[data["Age"].isna(), "Age"] = int(data["Age"].mode().to_list()[0])
        data = pd.get_dummies(data)
        # Get missing columns in the training test
        missing_cols = set( self.features ) - set( data.columns )
        # Add a missing column in test set with default value equal to 0
        for c in missing_cols:
            data[c] = 0
        # Ensure the order of column in the test set is in the same order than in train set
        # also, the columns which were not in training set(when fit was called) will be dropped.
        data = data[self.features]
        
        return data


columns = ["Pclass", "Sex", "Age", "SibSp", "Parch", "Fare"]
data = pd.read_csv("train.csv")[columns + ["Survived"]]

pipe = Pipeline([
        ("transform", DataTransformer()),
        ("classify", XGBClassifier())
        ])

y = data["Survived"]
X = data.drop(columns = ["Survived"])

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state = 0)

pipe.fit(X_train, y_train)

preds = pipe.predict(X_test)
print("-------XG Boost-------")
print("Accuracy is "+ str(metrics.accuracy_score(y_test, preds)))
print("Confusion Matrix is:")
print(metrics.confusion_matrix(y_test,preds))


# now train full training set for final test:
pipe.fit(X, y)

submissionX = pd.read_csv("test.csv")[columns]
submissiony = pipe.predict(submissionX)

final = pd.read_csv("test.csv")
final["Survived"] = submissiony

final[["PassengerId","Survived"]].to_csv("Submission.csv", index=False)