example.py

from __future__ import absolute_import, division, print_function, unicode_literals
import functools

import numpy as np
np.set_printoptions(precision=3, suppress=True)

import tensorflow as tf
import pandas as pd


def get_dataset(file_path, **kwargs):
    dataset = tf.data.experimental.make_csv_dataset(
      file_path,
      batch_size=5, # Artificially small to make examples easier to show.
      label_name='survived',
      na_value="?",
      num_epochs=1,
      ignore_errors=True, 
      **kwargs)
    return dataset


class PackNumericFeatures(object):
    def __init__(self, names):
        self.names = names

    def __call__(self, features, labels):
        numeric_features = [features.pop(name) for name in self.names]
        numeric_features = [tf.cast(feat, tf.float32) for feat in numeric_features]
        numeric_features = tf.stack(numeric_features, axis=-1)
        features['numeric'] = numeric_features

        return features, labels


def normalize_numeric_data(data, mean, std):
    # Center the data
    return (data-mean)/std


def show_batch(dataset):
    for batch, label in dataset.take(1):
        for key, value in batch.items():
            print("{:20s}: {}".format(key, value.numpy()))


# dataset
TRAIN_DATA_URL = "https://storage.googleapis.com/tf-datasets/titanic/train.csv"
TEST_DATA_URL = "https://storage.googleapis.com/tf-datasets/titanic/eval.csv"
train_file_path = tf.keras.utils.get_file("train.csv", TRAIN_DATA_URL)
test_file_path = tf.keras.utils.get_file("eval.csv", TEST_DATA_URL)
raw_train_data = get_dataset(train_file_path)
raw_test_data = get_dataset(test_file_path)



# number features preprocessing
NUMERIC_FEATURES = ['age','n_siblings_spouses','parch', 'fare']

packed_train_data = raw_train_data.map(
    PackNumericFeatures(NUMERIC_FEATURES))

packed_test_data = raw_test_data.map(
    PackNumericFeatures(NUMERIC_FEATURES))


desc = pd.read_csv(train_file_path)[NUMERIC_FEATURES].describe()
normalizer = functools.partial(normalize_numeric_data, mean=np.array(desc.T['mean']), std=np.array(desc.T['std']))
numeric_column = tf.feature_column.numeric_column('numeric', normalizer_fn=normalizer, shape=[len(NUMERIC_FEATURES)])
numeric_columns = [numeric_column]


# category features proprocessing
CATEGORIES = {
    'sex': ['male', 'female'],
    'class' : ['First', 'Second', 'Third'],
    'deck' : ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J'],
    'embark_town' : ['Cherbourg', 'Southhampton', 'Queenstown'],
    'alone' : ['y', 'n']
}

categorical_columns = []
for feature, vocab in CATEGORIES.items():
    cat_col = tf.feature_column.categorical_column_with_vocabulary_list(
        key=feature, vocabulary_list=vocab)
    categorical_columns.append(tf.feature_column.indicator_column(cat_col))


# model
model = tf.keras.Sequential([
  tf.keras.layers.DenseFeatures(categorical_columns + numeric_columns),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dense(1, activation='sigmoid'),
])

model.compile(
    loss='binary_crossentropy',
    optimizer='adam',
    metrics=['accuracy'])



# train, evaluate and predict
test_data = packed_test_data
model.fit(packed_train_data.shuffle(500), epochs=20)

test_loss, test_accuracy = model.evaluate(test_data)

print('\n\nTest Loss {}, Test Accuracy {}'.format(test_loss, test_accuracy))

predictions = model.predict(test_data)

# Show some results
for prediction, survived in zip(predictions[:10], list(test_data)[0][1][:10]):
    print("Predicted survival: {:.2%}".format(prediction[0]),
        " | Actual outcome: ",
        ("SURVIVED" if bool(survived) else "DIED"))