abstract PrepareData interface for example

README.md

@@ -72,6 +72,7 @@ Read the docs & book [The Definitive Guide to Tensorflow.NET](https://tensorflow
 * [Basic Operations](test/TensorFlowNET.Examples/BasicOperations.cs)
 * [Image Recognition](test/TensorFlowNET.Examples/ImageRecognition.cs)
 * [Linear Regression](test/TensorFlowNET.Examples/LinearRegression.cs)
+* [Logistic Regression](test/TensorFlowNET.Examples/LogisticRegression.cs)
 * [Text Classification](test/TensorFlowNET.Examples/TextClassificationWithMovieReviews.cs)
 * [CNN Text Classification](test/TensorFlowNET.Examples/CnnTextClassification.cs)
 * [Naive Bayes Classification](test/TensorFlowNET.Examples/NaiveBayesClassifier.cs)

src/TensorFlowNET.Core/TensorFlowNET.Core.csproj

@@ -17,7 +17,8 @@
     <Description>Google's TensorFlow binding in .NET Standard.
 Docs: https://tensorflownet.readthedocs.io</Description>
     <AssemblyVersion>0.5.0.0</AssemblyVersion>
-    <PackageReleaseNotes>Add a lot of APIs to build neural networks model</PackageReleaseNotes>
+    <PackageReleaseNotes>Add Logistic Regression to do MNIST.
+Add a lot of APIs to build neural networks model</PackageReleaseNotes>
     <LangVersion>7.2</LangVersion>
     <FileVersion>0.5.0.0</FileVersion>
   </PropertyGroup>

test/TensorFlowNET.Examples/BasicEagerApi.cs

@@ -12,6 +12,7 @@ namespace TensorFlowNET.Examples
     public class BasicEagerApi : IExample
     {
         private Tensor a, b, c, d;
+
         public void Run()
         {
             // Set Eager API
@@ -34,5 +35,9 @@ namespace TensorFlowNET.Examples
 
             // Full compatibility with Numpy
         }
+
+        public void PrepareData()
+        {
+        }
     }
 }

test/TensorFlowNET.Examples/BasicOperations.cs

@@ -96,5 +96,9 @@ namespace TensorFlowNET.Examples
                 }
             }
         }
+
+        public void PrepareData()
+        {
+        }
     }
 }

test/TensorFlowNET.Examples/HelloWorld.cs

@@ -33,5 +33,9 @@ namespace TensorFlowNET.Examples
                 }
             }
         }
+
+        public void PrepareData()
+        {
+        }
     }
 }

test/TensorFlowNET.Examples/IExample.cs

@@ -11,5 +11,6 @@ namespace TensorFlowNET.Examples
     public interface IExample
     {
         void Run();
+        void PrepareData();
     }
 }

test/TensorFlowNET.Examples/ImageRecognition.cs

@@ -78,7 +78,7 @@ namespace TensorFlowNET.Examples
             });
         }
 
-        private void PrepareData()
+        public void PrepareData()
         {
             Directory.CreateDirectory(dir);
 

test/TensorFlowNET.Examples/InceptionArchGoogLeNet.cs

@@ -83,7 +83,7 @@ namespace TensorFlowNET.Examples
             });
         }
 
-        private void PrepareData()
+        public void PrepareData()
         {
             Directory.CreateDirectory(dir);
 

test/TensorFlowNET.Examples/LinearRegression.cs

@@ -19,14 +19,13 @@ namespace TensorFlowNET.Examples
         int training_epochs = 1000;
         int display_step = 50;
 
+        NDArray train_X, train_Y;
+        int n_samples;
+
         public void Run()
         {
             // Training Data
-            var train_X = np.array(3.3f, 4.4f, 5.5f, 6.71f, 6.93f, 4.168f, 9.779f, 6.182f, 7.59f, 2.167f,
-                         7.042f, 10.791f, 5.313f, 7.997f, 5.654f, 9.27f, 3.1f);
-            var train_Y = np.array(1.7f, 2.76f, 2.09f, 3.19f, 1.694f, 1.573f, 3.366f, 2.596f, 2.53f, 1.221f,
-                         2.827f, 3.465f, 1.65f, 2.904f, 2.42f, 2.94f, 1.3f);
-            var n_samples = train_X.shape[0];
+            PrepareData();
 
             // tf Graph Input
             var X = tf.placeholder(tf.float32);
@@ -95,5 +94,14 @@ namespace TensorFlowNET.Examples
                 Console.WriteLine($"Absolute mean square loss difference: {Math.Abs((float)training_cost - (float)testing_cost)}");
             });
         }
+
+        public void PrepareData()
+        {
+            train_X = np.array(3.3f, 4.4f, 5.5f, 6.71f, 6.93f, 4.168f, 9.779f, 6.182f, 7.59f, 2.167f,
+             7.042f, 10.791f, 5.313f, 7.997f, 5.654f, 9.27f, 3.1f);
+            train_Y = np.array(1.7f, 2.76f, 2.09f, 3.19f, 1.694f, 1.573f, 3.366f, 2.596f, 2.53f, 1.221f,
+                         2.827f, 3.465f, 1.65f, 2.904f, 2.42f, 2.94f, 1.3f);
+            n_samples = train_X.shape[0];
+        }
     }
 }

test/TensorFlowNET.Examples/LogisticRegression.cs

@@ -21,9 +21,11 @@ namespace TensorFlowNET.Examples
         private int batch_size = 100;
         private int display_step = 1;
 
+        Datasets mnist;
+
         public void Run()
         {
-            var mnist = PrepareData();
+            PrepareData();
 
             // tf Graph Input
             var x = tf.placeholder(tf.float32, new TensorShape(-1, 784)); // mnist data image of shape 28*28=784
@@ -86,10 +88,9 @@ namespace TensorFlowNET.Examples
             });
         }
 
-        private Datasets PrepareData()
+        public void PrepareData()
         {
-            var mnist = MnistDataSet.read_data_sets("logistic_regression", one_hot: true);
-            return mnist;
+            mnist = MnistDataSet.read_data_sets("logistic_regression", one_hot: true);
         }
     }
 }

test/TensorFlowNET.Examples/MetaGraph.cs

@@ -27,5 +27,9 @@ namespace TensorFlowNET.Examples
                                                 logits: logits);
             });
         }
+
+        public void PrepareData()
+        {
+        }
     }
 }

test/TensorFlowNET.Examples/NaiveBayesClassifier.cs

@@ -15,9 +15,9 @@ namespace TensorFlowNET.Examples
         public Normal dist { get; set; }
         public void Run()
         {
-            np.array<float>(1.0f, 1.0f);
-            var X = np.array<float>(new float[][] { new float[] { 1.0f, 1.0f }, new float[] { 2.0f, 2.0f }, new float[] { -1.0f, -1.0f }, new float[] { -2.0f, -2.0f }, new float[] { 1.0f, -1.0f }, new float[] { 2.0f, -2.0f }, });
-            var y = np.array<int>(0,0,1,1,2,2);
+            np.array(1.0f, 1.0f);
+            var X = np.array(new float[][] { new float[] { 1.0f, 1.0f }, new float[] { 2.0f, 2.0f }, new float[] { -1.0f, -1.0f }, new float[] { -2.0f, -2.0f }, new float[] { 1.0f, -1.0f }, new float[] { 2.0f, -2.0f }, });
+            var y = np.array(0,0,1,1,2,2);
             fit(X, y);
             // Create a regular grid and classify each point 
         }
@@ -102,5 +102,10 @@ namespace TensorFlowNET.Examples
             // exp to get the actual probabilities
             return tf.exp(log_prob);
         }
+
+        public void PrepareData()
+        {
+            
+        }
     }
 }

test/TensorFlowNET.Examples/NamedEntityRecognition.cs

@@ -14,5 +14,10 @@ namespace TensorFlowNET.Examples
         {
             throw new NotImplementedException();
         }
+
+        public void PrepareData()
+        {
+            throw new NotImplementedException();
+        }
     }
 }

test/TensorFlowNET.Examples/TextClassification/TextClassificationTrain.cs

@@ -23,7 +23,7 @@ namespace TensorFlowNET.Examples.CnnTextClassification
 
         public void Run()
         {
-            download_dbpedia();
+            PrepareData();
             Console.WriteLine("Building dataset...");
             var (x, y, alphabet_size) = DataHelpers.build_char_dataset("train", "vdcnn", CHAR_MAX_LEN);
 
@@ -32,17 +32,9 @@ namespace TensorFlowNET.Examples.CnnTextClassification
             with(tf.Session(), sess =>
             {
                 new VdCnn(alphabet_size, CHAR_MAX_LEN, NUM_CLASS);
-
             });
         }
 
-        public void download_dbpedia()
-        {
-            string url = "https://github.com/le-scientifique/torchDatasets/raw/master/dbpedia_csv.tar.gz";
-            Web.Download(url, dataDir, dataFileName);
-            Compress.ExtractTGZ(Path.Join(dataDir, dataFileName), dataDir);
-        }
-
         private (int[][], int[][], int[], int[]) train_test_split(int[][] x, int[] y, float test_size = 0.3f)
         {
             int len = x.Length;
@@ -75,5 +67,12 @@ namespace TensorFlowNET.Examples.CnnTextClassification
 
             return (train_x.ToArray(), valid_x.ToArray(), train_y.ToArray(), valid_y.ToArray());
         }
+
+        public void PrepareData()
+        {
+            string url = "https://github.com/le-scientifique/torchDatasets/raw/master/dbpedia_csv.tar.gz";
+            Web.Download(url, dataDir, dataFileName);
+            Compress.ExtractTGZ(Path.Join(dataDir, dataFileName), dataDir);
+        }
     }
 }

test/TensorFlowNET.Examples/TextClassification/TextClassificationWithMovieReviews.cs

@@ -13,10 +13,11 @@ namespace TensorFlowNET.Examples
     {
         string dir = "text_classification_with_movie_reviews";
         string dataFile = "imdb.zip";
+        NDArray train_data, train_labels, test_data, test_labels;
 
         public void Run()
         {
-            var((train_data, train_labels), (test_data, test_labels)) = PrepareData();
+            PrepareData();
 
             Console.WriteLine($"Training entries: {train_data.size}, labels: {train_labels.size}");
 
@@ -40,7 +41,7 @@ namespace TensorFlowNET.Examples
             model.add(keras.layers.Embedding(vocab_size, 16));
         }
 
-        private ((NDArray, NDArray), (NDArray, NDArray)) PrepareData()
+        public void PrepareData()
         {
             Directory.CreateDirectory(dir);
 
@@ -71,7 +72,11 @@ namespace TensorFlowNET.Examples
             var y_train = labels_train;
             var y_test = labels_test;
 
-            return ((x_train, y_train), (x_test, y_test));
+            x_train = train_data;
+            train_labels = y_train;
+
+            test_data = x_test;
+            test_labels = y_test;
         }
 
         private NDArray ReadData(string file)