I am trying to construct a dataset to use in Keras for the Titanic example on Kaggle. Here is what I\'ve done so far:
train_data = pd.read_csv("/kaggle/input