Merge branch 'master' into neural-translation

Author: ilblackdragon

.coveragerc

@@ -2,5 +2,6 @@
 branch = True
 
 [report]
-omit = */tests/*
-
+omit = 
+        */tests/*
+        */ops/batch_norm_ops.py

.pylintrc

@@ -2,7 +2,7 @@
 
 ignore=tests
 
-disable=invalid-name,wildcard-import,too-many-arguments,attribute-defined-outside-init,no-member,too-many-instance-attributes,too-few-public-methods,import-error,super-on-old-class,fixme,protected-access,locally-disabled,cyclic-import,too-many-locals
+disable=invalid-name,wildcard-import,too-many-arguments,attribute-defined-outside-init,no-member,too-many-instance-attributes,too-few-public-methods,import-error,super-on-old-class,fixme,protected-access,locally-disabled,cyclic-import,too-many-locals,duplicate-code
 
 dummy-variables-rgx=(unused|)_.*
 

.travis.yml

@@ -19,7 +19,6 @@ cache:
 addons:
     apt:
         packages:
-            - libatlas3gf-base
             - libatlas-dev
             - libblas-dev
             - liblapack-dev

README.md

@@ -1,17 +1,19 @@
-[![Travis-CI Build Status](https://travis-ci.org/google/skflow.svg?branch=master)](https://travis-ci.org/google/skflow)
-[![Codecov Status](https://codecov.io/github/google/skflow/coverage.svg?precision=2)](https://codecov.io/github/google/skflow)
-[![License](https://img.shields.io/github/license/google/skflow.svg)](http://www.apache.org/licenses/LICENSE-2.0.html)
-[![Join the chat at https://gitter.im/google/skflow](https://badges.gitter.im/Join%20Chat.svg)](https://gitter.im/google/skflow?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
+[![Travis-CI Build Status](https://travis-ci.org/tensorflow/skflow.svg?branch=master)](https://travis-ci.org/tensorflow/skflow)
+[![Codecov Status](https://codecov.io/github/tensorflow/skflow/coverage.svg?precision=2)](https://codecov.io/github/tensorflow/skflow)
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](http://www.apache.org/licenses/LICENSE-2.0.html)
+[![Join the chat at https://gitter.im/tensorflow/skflow](https://badges.gitter.im/Join%20Chat.svg)](https://gitter.im/tensorflow/skflow?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
 
 # Scikit Flow
 
 This is a simplified interface for TensorFlow, to get people started on predictive analytics and data mining.
 
-Why TensorFlow?
+Library covers variety of needs from linear models to *Deep Learning* applications like text and image understanding.
+
+Why *TensorFlow*?
 * TensorFlow provides a good backbone for building different shapes of machine learning applications.
 * It will continue to evolve both in the distributed direction and as general pipelinining machinery.
 
-Why Scikit Flow?
+Why *Scikit Flow*?
 * To smooth the transition from the Scikit Learn world of one-liner machine learning into the
 more open world of building different shapes of ML models. You can start by using fit/predict and slide into TensorFlow APIs as you are getting comfortable.
 * To provide a set of reference models that would be easy to integrate with existing code.
@@ -26,19 +28,20 @@ Support versions of dependencies:
 First, make sure you have TensorFlow and Scikit Learn installed, then just run:
 
 ```Bash
-pip install git+git://github.com/google/skflow.git
+pip install git+git://github.com/tensorflow/skflow.git
 ```
 
 ## Tutorial
 
 * [Introduction to Scikit Flow and why you want to start learning TensorFlow](https://medium.com/@ilblackdragon/tensorflow-tutorial-part-1-c559c63c0cb1)
 * [DNNs, custom model and Digit recognition examples](https://medium.com/@ilblackdragon/tensorflow-tutorial-part-2-9ffe47049c92)
+* [Categorical variables: One hot vs Distributed representation](https://medium.com/@ilblackdragon/tensorflow-tutorial-part-3-c5fc0662bc08)
 * More coming soon.
 
 ## Usage
 
 Below are few simple examples of the API. 
-For more examples, please see [examples](https://github.com/google/skflow/tree/master/examples).
+For more examples, please see [examples](https://github.com/tensorflow/skflow/tree/master/examples).
 
 ### General tips
 
@@ -58,7 +61,7 @@ from sklearn import datasets, metrics
 iris = datasets.load_iris()
 classifier = skflow.TensorFlowLinearClassifier(n_classes=3)
 classifier.fit(iris.data, iris.target)
-score = metrics.accuracy_score(classifier.predict(iris.data), iris.target)
+score = metrics.accuracy_score(iris.target, classifier.predict(iris.data))
 print("Accuracy: %f" % score)
 ```
 
@@ -89,7 +92,7 @@ from sklearn import datasets, metrics
 iris = datasets.load_iris()
 classifier = skflow.TensorFlowDNNClassifier(hidden_units=[10, 20, 10], n_classes=3)
 classifier.fit(iris.data, iris.target)
-score = metrics.accuracy_score(classifier.predict(iris.data), iris.target)
+score = metrics.accuracy_score(iris.target, classifier.predict(iris.data))
 print("Accuracy: %f" % score)
 ```
 
@@ -110,7 +113,7 @@ def my_model(X, y):
 
 classifier = skflow.TensorFlowEstimator(model_fn=my_model, n_classes=3)
 classifier.fit(iris.data, iris.target)
-score = metrics.accuracy_score(classifier.predict(iris.data), iris.target)
+score = metrics.accuracy_score(iris.target, classifier.predict(iris.data))
 print("Accuracy: %f" % score)
 ```
 
@@ -172,10 +175,10 @@ tensorboard --logdir=/tmp/tf_examples/my_model_1
 and follow reported url.
 
 Graph visualization:
-![Text classification RNN Graph](https://raw.githubusercontent.com/google/skflow/master/docs/images/text_classification_rnn_graph.png)
+![Text classification RNN Graph](https://raw.githubusercontent.com/tensorflow/skflow/master/docs/images/text_classification_rnn_graph.png)
 
 Loss visualization:
-![Text classification RNN Loss](https://raw.githubusercontent.com/google/skflow/master/docs/images/text_classification_rnn_loss.png)
+![Text classification RNN Loss](https://raw.githubusercontent.com/tensorflow/skflow/master/docs/images/text_classification_rnn_loss.png)
 
 ## More examples
 

examples/README.md

@@ -4,18 +4,28 @@ Examples of Using skflow
 * [Deep Neural Network Regression with Boston Data](boston.py)
 * [Convolutional Neural Networks with Digits Data](digits.py)
 * [Deep Neural Network Classification with Iris Data](iris.py)
+* [Deep Neural Network with Customized Decay Function](iris_custom_decay_dnn.py)
 * [Building A Custom Model](iris_custom_model.py)
 * [Building A Custom Model Using Multiple GPUs](multiple_gpu.py)
 * [Using skflow with Pipeline](iris_with_pipeline.py)
 * [Example of saving and restoring models](iris_save_restore.py)
 * [Multi-output Deep Neural Network regression](multioutput_regression.py)
 * [Improving Performance Using Early Stopping with Iris Data](iris_early_stopping.py)
+* [Out-of-core Data Classification Using Dask](out_of_core_data_classification.py)
+
+
+Image classification
+--------------------
+
+* [Convolutional Neural Networks on MNIST Data](mnist.py)
+* [Deep Residual Networks on MNIST Data](resnet.py)
 
 
 Text classification
 -------------------
 
-* [Text Classification Using Recurrent Neural Networks on Words](text_classification.py)
+* [Text Classification Using Recurrent Neural Networks on Words](text_classification.py) (See also [Simplified Version Using Built-in RNN Model](text_classification_builtin_rnn_model.py))
 * [Text Classification Using Convolutional Neural Networks on Words](text_classification_cnn.py)
 * [Text Classification Using Recurrent Neural Networks on Characters](text_classification_character_rnn.py)
 * [Text Classification Using Convolutional Neural Networks on Characters](text_classification_character_cnn.py)
+

examples/boston.py

@@ -25,19 +25,22 @@
 boston = datasets.load_boston()
 X, y = boston.data, boston.target
 
-# Scale data to 0 mean and unit std dev.
-scaler = preprocessing.StandardScaler()
-X = scaler.fit_transform(X)
-
 # Split dataset into train / test
 X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y,
     test_size=0.2, random_state=42)
 
+# scale data (training set) to 0 mean and unit Std. dev
+scaler = preprocessing.StandardScaler()
+X_train = scaler.fit_transform(X_train)
+
 # Build 2 layer fully connected DNN with 10, 10 units respecitvely.
 regressor = skflow.TensorFlowDNNRegressor(hidden_units=[10, 10],
     steps=5000, learning_rate=0.1, batch_size=1)
 
-# Fit and predict.
+# Fit
 regressor.fit(X_train, y_train)
-score = metrics.mean_squared_error(regressor.predict(X_test), y_test)
+
+# Predict and score
+score = metrics.mean_squared_error(regressor.predict(scaler.fit_transform(X_test)), y_test)
+
 print('MSE: {0:f}'.format(score))

examples/digits.py

@@ -18,8 +18,6 @@
 
 import skflow
 
-random.seed(42)
-
 # Load dataset and split it into train / test subsets.
 
 digits = datasets.load_digits()
@@ -42,5 +40,5 @@ def conv_model(X, y):
                                         steps=500, learning_rate=0.05,
                                         batch_size=128)
 classifier.fit(X_train, y_train)
-score = metrics.accuracy_score(classifier.predict(X_test), y_test)
+score = metrics.accuracy_score(y_test, classifier.predict(X_test))
 print('Accuracy: {0:f}'.format(score))

examples/iris.py

@@ -27,10 +27,10 @@
 
 # Build 3 layer DNN with 10, 20, 10 units respecitvely.
 classifier = skflow.TensorFlowDNNClassifier(hidden_units=[10, 20, 10],
-n_classes=3, steps=200)
+    n_classes=3, steps=200)
 
 # Fit and predict.
 classifier.fit(X_train, y_train)
-score = metrics.accuracy_score(classifier.predict(X_test), y_test)
+score = metrics.accuracy_score(y_test, classifier.predict(X_test))
 print('Accuracy: {0:f}'.format(score))
 

examples/iris_custom_decay_dnn.py

@@ -0,0 +1,42 @@
+#  Copyright 2015-present Scikit Flow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+
+import random
+
+from sklearn import datasets, metrics
+from sklearn.cross_validation import train_test_split
+
+import skflow
+import tensorflow as tf
+
+
+random.seed(42)
+
+iris = datasets.load_iris()
+X_train, X_test, y_train, y_test = train_test_split(iris.data,
+                                                    iris.target,
+                                                    test_size=0.2,
+                                                    random_state=42)
+# setup exponential decay function
+def exp_decay(global_step):
+    return tf.train.exponential_decay(
+        learning_rate=0.1, global_step=global_step,
+        decay_steps=100, decay_rate=0.001)
+
+# use customized decay function in learning_rate
+classifier = skflow.TensorFlowDNNClassifier(hidden_units=[10, 20, 10],
+                                            n_classes=3, steps=800,
+                                            learning_rate=exp_decay)
+classifier.fit(X_train, y_train)
+score = metrics.accuracy_score(y_test, classifier.predict(X_test))

examples/iris_custom_model.py

@@ -24,12 +24,13 @@
 random.seed(42)
 
 def my_model(X, y):
-    """This is DNN with 10, 20, 10 hidden layers, and dropout of 0.5 probability."""
-    layers = skflow.ops.dnn(X, [10, 20, 10], keep_prob=0.5)
+    """This is DNN with 10, 20, 10 hidden layers, and dropout of 0.9 probability."""
+    layers = skflow.ops.dnn(X, [10, 20, 10], keep_prob=0.9)
     return skflow.models.logistic_regression(layers, y)
 
-classifier = skflow.TensorFlowEstimator(model_fn=my_model, n_classes=3)
+classifier = skflow.TensorFlowEstimator(model_fn=my_model, n_classes=3,
+    steps=1000)
 classifier.fit(X_train, y_train)
-score = metrics.accuracy_score(classifier.predict(X_test), y_test)
+score = metrics.accuracy_score(y_test, classifier.predict(X_test))
 print('Accuracy: {0:f}'.format(score))