Created
July 12, 2018 23:17
-
-
Save securetorobert/d248d4c359a6b5d20941aa5502d135b3 to your computer and use it in GitHub Desktop.
Linear Regression with TensorFlow canned estimators
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import numpy as np | |
import pandas as pd | |
import tensorflow as tf | |
from sklearn.model_selection import train_test_split | |
train_df = pd.read_csv('./train.csv') | |
#create feature columns | |
crim = tf.feature_column.numeric_column('crim', dtype=tf.float64, shape=()) | |
zn = tf.feature_column.numeric_column('zn', dtype=tf.float64, shape=()) | |
indus = tf.feature_column.numeric_column('indus', dtype=tf.float64, shape=()) | |
chas = tf.feature_column.numeric_column('chas', dtype=tf.int64, shape=()) | |
nox = tf.feature_column.numeric_column('nox', dtype=tf.float64, shape=()) | |
rm = tf.feature_column.numeric_column('rm', dtype=tf.float64, shape=()) | |
age = tf.feature_column.numeric_column('age', dtype=tf.float64, shape=()) | |
dis = tf.feature_column.numeric_column('dis', dtype=tf.float64, shape=()) | |
rad = tf.feature_column.numeric_column('rad', dtype=tf.int64, shape=()) | |
tax = tf.feature_column.numeric_column('tax', dtype=tf.int64, shape=()) | |
ptratio = tf.feature_column.numeric_column('ptratio', dtype=tf.float64, shape=()) | |
black = tf.feature_column.numeric_column('black', dtype=tf.float64, shape=()) | |
lstat = tf.feature_column.numeric_column('lstat', dtype=tf.float64, shape=()) | |
feature_cols = [crim, zn, indus, chas, nox, rm, age, dis, rad, tax, ptratio, black, lstat] | |
#create training and validation data sets | |
feature_names = ['crim', 'zn', 'indus', 'chas', 'nox', 'rm', 'age', 'dis', 'rad', 'tax', 'ptratio', 'black', 'lstat'] | |
label_name = 'medv' | |
features_ndarray = train_df[feature_names] | |
label_ndarray = train_df[label_name] | |
X_train, X_test, y_train, y_test = train_test_split(features_ndarray, label_ndarray, random_state=0, test_size=0.3) | |
#create input functions | |
def train_input(): | |
_dataset = tf.data.Dataset.from_tensor_slices(({'crim': X_train['crim'], | |
'zn': X_train['zn'], | |
'indus': X_train['indus'], | |
'chas': X_train['chas'], | |
'nox': X_train['nox'], | |
'rm': X_train['rm'], | |
'age': X_train['age'], | |
'dis': X_train['dis'], | |
'rad': X_train['rad'], | |
'tax': X_train['tax'], | |
'ptratio': X_train['ptratio'], | |
'black': X_train['black'], | |
'lstat': X_train['lstat'] | |
}, y_train)) | |
dataset = _dataset.batch(32) | |
iterator = dataset.make_one_shot_iterator() | |
features, labels = iterator.get_next() | |
return features, labels | |
def val_input(): | |
_dataset = tf.data.Dataset.from_tensor_slices(({'crim': X_test['crim'], | |
'zn': X_test['zn'], | |
'indus': X_test['indus'], | |
'chas': X_test['chas'], | |
'nox': X_test['nox'], | |
'rm': X_test['rm'], | |
'age': X_test['age'], | |
'dis': X_test['dis'], | |
'rad': X_test['rad'], | |
'tax': X_test['tax'], | |
'ptratio': X_test['ptratio'], | |
'black': X_test['black'], | |
'lstat': X_test['lstat'] | |
}, y_test)) | |
dataset = _dataset.batch(32) | |
iterator = dataset.make_one_shot_iterator() | |
features, labels = iterator.get_next() | |
return features, labels | |
#instantiate our estimator | |
estimator = tf.estimator.LinearRegressor(feature_columns=feature_cols) | |
#train our estimator | |
estimator.train(input_fn=train_input, steps=None) | |
#evaluation | |
train_e = estimator.evaluate(input_fn=train_input) | |
test_e = estimator.evaluate(input_fn=val_input) | |
#inference | |
preds = estimator.predict(input_fn=val_input) | |
predictions = np.array([item['predictions'][0] for item in preds]) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment