using System.Collections.Generic;
using Tensorflow;
using Tensorflow.Keras;
using static Tensorflow.Binding;
using static Tensorflow.KerasApi;
using Tensorflow.Keras.Utils;
using System.IO;
using Tensorflow.Keras.Engine;

namespace TensorFlowNET.Examples
{
 /// <summary>
 /// This tutorial shows how to classify images of flowers.
 /// https://www.tensorflow.org/tutorials/images/classification
 /// </summary>
 public class ImageClassificationKeras : SciSharpExample, IExample
 {
 int batch_size = 32;
 int epochs = 3;
 TensorShape img_dim = (180, 180);
 IDatasetV2 train_ds, val_ds;
 Model model;

 public ExampleConfig InitConfig()
 => Config = new ExampleConfig
 {
 Name = "Image Classification (Keras)",
 Enabled = true,
 Priority = 18
 };

 public bool Run()
 {
 tf.enable_eager_execution();

 PrepareData();
 BuildModel();
 Train();

 return true;
 }

 public override void BuildModel()
 {
 int num_classes = 5;
 // var normalization_layer = tf.keras.layers.Rescaling(1.0f / 255);
 var layers = keras.layers;
 model = keras.Sequential(new List<ILayer>
 {
 layers.Rescaling(1.0f / 255, input_shape: (img_dim.dims[0], img_dim.dims[1], 3)),
 layers.Conv2D(16, 3, padding: "same", activation: keras.activations.Relu),
 layers.MaxPooling2D(),
 /*layers.Conv2D(32, 3, padding: "same", activation: "relu"),
 layers.MaxPooling2D(),
 layers.Conv2D(64, 3, padding: "same", activation: "relu"),
 layers.MaxPooling2D(),*/
 layers.Flatten(),
 layers.Dense(128, activation: keras.activations.Relu),
 layers.Dense(num_classes)
 });

 model.compile(optimizer: keras.optimizers.Adam(),
 loss: keras.losses.SparseCategoricalCrossentropy(from_logits: true),
 metrics: new[] { "accuracy" });

 model.summary();
 }

 public override void Train()
 {
 model.fit(train_ds, validation_data: val_ds, epochs: epochs);
 }

 public override void PrepareData()
 {
 string fileName = "flower_photos.tgz";
 string url = $"https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz";
 string data_dir = Path.Combine(Path.GetTempPath(), "flower_photos");
 Web.Download(url, data_dir, fileName);
 Compress.ExtractTGZ(Path.Join(data_dir, fileName), data_dir);
 data_dir = Path.Combine(data_dir, "flower_photos");

 // convert to tensor
 train_ds = keras.preprocessing.image_dataset_from_directory(data_dir,
 validation_split: 0.2f,
 subset: "training",
 seed: 123,
 image_size: img_dim,
 batch_size: batch_size);

 val_ds = keras.preprocessing.image_dataset_from_directory(data_dir,
 validation_split: 0.2f,
 subset: "validation",
 seed: 123,
 image_size: img_dim,
 batch_size: batch_size);

 train_ds = train_ds.shuffle(1000).prefetch(buffer_size: -1);
 val_ds = val_ds.prefetch(buffer_size: -1);

 foreach (var (img, label) in train_ds)
 {
 print($"images: {img.TensorShape}");
 print($"labels: {label.numpy()}");
 }
 }
 }
}