{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import math, re, os, random\nimport tensorflow as tf\nimport numpy as np\nfrom matplotlib import pyplot as plt\n#from kaggle_datasets import KaggleDatasets\nfrom sklearn.metrics import f1_score, precision_score, recall_score, confusion_matrix\nprint(\"Tensorflow version \" + tf.__version__)\nAUTO = tf.data.experimental.AUTOTUNE","metadata":{"id":"nF5BZRpzx8cz","outputId":"346656a4-6b26-418a-c04f-6c3466a42af3","execution":{"iopub.status.busy":"2021-12-28T11:51:36.722641Z","iopub.execute_input":"2021-12-28T11:51:36.723466Z","iopub.status.idle":"2021-12-28T11:51:43.261266Z","shell.execute_reply.started":"2021-12-28T11:51:36.723361Z","shell.execute_reply":"2021-12-28T11:51:43.260391Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Detect hardware, return appropriate distribution strategy\ntry:\n    tpu = tf.distribute.cluster_resolver.TPUClusterResolver()  # TPU detection. No parameters necessary if TPU_NAME environment variable is set. On Kaggle this is always the case.\n    print('Running on TPU ', tpu.master())\nexcept ValueError:\n    tpu = None\n\nif tpu:\n    tf.config.experimental_connect_to_cluster(tpu)\n    tf.tpu.experimental.initialize_tpu_system(tpu)\n    strategy = tf.distribute.experimental.TPUStrategy(tpu)\nelse:\n    strategy = tf.distribute.get_strategy() # default distribution strategy in Tensorflow. Works on CPU and single GPU.\n\nprint(\"REPLICAS: \", strategy.num_replicas_in_sync)","metadata":{"id":"ZortilVjyB3t","outputId":"5435c064-45d7-4218-fb01-fd1b1b97aa28","execution":{"iopub.status.busy":"2021-12-28T11:51:43.262701Z","iopub.execute_input":"2021-12-28T11:51:43.262924Z","iopub.status.idle":"2021-12-28T11:51:49.866911Z","shell.execute_reply.started":"2021-12-28T11:51:43.262898Z","shell.execute_reply":"2021-12-28T11:51:49.866014Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"code","source":"#GCS_DS_PATH = \"gs://kds-e4e51a1bf1ae2e1621131f2ccb169a42cdae2fe9b516885c29f2357f\"\n#print(GCS_DS_PATH)\n#PATH2 = \"gs://kds-4a5e3a920317029e41071344762bc9a9368b35492f7ae4f60649d\"\n#PATH2=KaggleDatasets().get_gcs_path('tf-flower-photo-tfrec')\n#from kaggle_datasets import KaggleDatasets\n#GCS_DS_PATH = KaggleDatasets().get_gcs_path('tpu-getting-started')\n#print(GCS_DS_PATH) \n# what do gcs paths look like?\n#GCS_DS_PATH = \"gs://kds-6d8f09dbd2512db4b94e7063db16ed25cc9d5ee315207d6971dd7644\"\n#PATH2 = KaggleDatasets().get_gcs_path('tf-flower-photo-tfrec')\n#print(PATH2)\n#print(PATH2)\nfrom kaggle_datasets import KaggleDatasets\nGCS_DS_PATH = KaggleDatasets().get_gcs_path('tpu-getting-started')\nprint(GCS_DS_PATH) # what do gcs paths look like?\n#GCS_DS_PATH = \"gs://kds-6d8f09dbd2512db4b94e7063db16ed25cc9d5ee315207d6971dd7644\"\nPATH2 = KaggleDatasets().get_gcs_path('tf-flower-photo-tfrec')\nprint(PATH2)","metadata":{"id":"YJD1LVOdyF8x","outputId":"113ac7c9-87a7-466d-d35c-f90425b03478","execution":{"iopub.status.busy":"2021-12-28T12:49:48.267222Z","iopub.execute_input":"2021-12-28T12:49:48.267523Z","iopub.status.idle":"2021-12-28T12:49:49.291386Z","shell.execute_reply.started":"2021-12-28T12:49:48.267494Z","shell.execute_reply":"2021-12-28T12:49:49.29069Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"IMAGE_SIZE = [512, 512] # At this size, a GPU will run out of memory. Use the TPU.\n                        # For GPU training, please select 224 x 224 px image size.\nEPOCHS = 12\nBATCH_SIZE =16 * strategy.num_replicas_in_sync\nimg_size=512\nGCS_PATH_SELECT = { # available image sizes\n    192: GCS_DS_PATH + '/tfrecords-jpeg-192x192',\n    224: GCS_DS_PATH + '/tfrecords-jpeg-224x224',\n    331: GCS_DS_PATH + '/tfrecords-jpeg-331x331',\n    512: GCS_DS_PATH + '/tfrecords-jpeg-512x512'\n}\nGCS_PATH = GCS_PATH_SELECT[IMAGE_SIZE[0]]\n\nTRAINING_FILENAMES = tf.io.gfile.glob(GCS_PATH + '/train/*.tfrec')\nTRAINING_FILENAMES += tf.io.gfile.glob(PATH2 + '/imagenet/tfrecords-jpeg-512x512/*.tfrec')\nTRAINING_FILENAMES += tf.io.gfile.glob(PATH2 + '/inaturalist_1/tfrecords-jpeg-512x512/*.tfrec')\nTRAINING_FILENAMES += tf.io.gfile.glob(PATH2 + '/openimage/tfrecords-jpeg-512x512/*.tfrec')\nTRAINING_FILENAMES += tf.io.gfile.glob(PATH2 + '/oxford_102/tfrecords-jpeg-512x512/*.tfrec')\nTRAINING_FILENAMES += tf.io.gfile.glob(PATH2 + '/tf_flowers/tfrecords-jpeg-512x512/*.tfrec')\nVALIDATION_FILENAMES = tf.io.gfile.glob(GCS_PATH + '/val/*.tfrec')\nTEST_FILENAMES = tf.io.gfile.glob(GCS_PATH + '/test/*.tfrec') # predictions on this dataset should be submitted for the competition\n#TRAINING_FILENAMES=TRAINING_FILENAMES+VALIDATION_FILENAMES\n\nCLASSES = ['pink primrose',    'hard-leaved pocket orchid', 'canterbury bells', 'sweet pea',     'wild geranium',     'tiger lily',           'moon orchid',              'bird of paradise', 'monkshood',        'globe thistle',         # 00 - 09\n           'snapdragon',       \"colt's foot\",               'king protea',      'spear thistle', 'yellow iris',       'globe-flower',         'purple coneflower',        'peruvian lily',    'balloon flower',   'giant white arum lily', # 10 - 19\n           'fire lily',        'pincushion flower',         'fritillary',       'red ginger',    'grape hyacinth',    'corn poppy',           'prince of wales feathers', 'stemless gentian', 'artichoke',        'sweet william',         # 20 - 29\n           'carnation',        'garden phlox',              'love in the mist', 'cosmos',        'alpine sea holly',  'ruby-lipped cattleya', 'cape flower',              'great masterwort', 'siam tulip',       'lenten rose',           # 30 - 39\n           'barberton daisy',  'daffodil',                  'sword lily',       'poinsettia',    'bolero deep blue',  'wallflower',           'marigold',                 'buttercup',        'daisy',            'common dandelion',      # 40 - 49\n           'petunia',          'wild pansy',                'primula',          'sunflower',     'lilac hibiscus',    'bishop of llandaff',   'gaura',                    'geranium',         'orange dahlia',    'pink-yellow dahlia',    # 50 - 59\n           'cautleya spicata', 'japanese anemone',          'black-eyed susan', 'silverbush',    'californian poppy', 'osteospermum',         'spring crocus',            'iris',             'windflower',       'tree poppy',            # 60 - 69\n           'gazania',          'azalea',                    'water lily',       'rose',          'thorn apple',       'morning glory',        'passion flower',           'lotus',            'toad lily',        'anthurium',             # 70 - 79\n           'frangipani',       'clematis',                  'hibiscus',         'columbine',     'desert-rose',       'tree mallow',          'magnolia',                 'cyclamen ',        'watercress',       'canna lily',            # 80 - 89\n           'hippeastrum ',     'bee balm',                  'pink quill',       'foxglove',      'bougainvillea',     'camellia',             'mallow',                   'mexican petunia',  'bromelia',         'blanket flower',        # 90 - 99\n           'trumpet creeper',  'blackberry lily',           'common tulip',     'wild rose']","metadata":{"id":"xs3v2s0IyIkf","execution":{"iopub.status.busy":"2021-12-28T12:49:54.942641Z","iopub.execute_input":"2021-12-28T12:49:54.942931Z","iopub.status.idle":"2021-12-28T12:49:56.227389Z","shell.execute_reply.started":"2021-12-28T12:49:54.9429Z","shell.execute_reply":"2021-12-28T12:49:56.22653Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# numpy and matplotlib defaults\nnp.set_printoptions(threshold=15, linewidth=80)\n\ndef batch_to_numpy_images_and_labels(data):\n    images, labels = data\n    numpy_images = images.numpy()\n    numpy_\n    if numpy_labels.dtype == object: # binary stringlabels = labels.numpy() in this case, these are image ID strings\n        numpy_labels = [None for _ in enumerate(numpy_images)]\n    # If no labels, only image IDs, return None for labels (this is the case for test data)\n    return numpy_images, numpy_labels\n\ndef title_from_label_and_target(label, correct_label):\n    if correct_label is None:\n        return CLASSES[label], True\n    correct = (label == correct_label)\n    return \"{} [{}{}{}]\".format(CLASSES[label], 'OK' if correct else 'NO', u\"\\u2192\" if not correct else '',\n                                CLASSES[correct_label] if not correct else ''), correct\n\ndef display_one_flower(image, title, subplot, red=False, titlesize=16):\n    plt.subplot(*subplot)\n    plt.axis('off')\n    plt.imshow(image)\n    if len(title) > 0:\n        plt.title(title, fontsize=int(titlesize) if not red else int(titlesize/1.2), color='red' if red else 'black', fontdict={'verticalalignment':'center'}, pad=int(titlesize/1.5))\n    return (subplot[0], subplot[1], subplot[2]+1)\n    \ndef display_batch_of_images(databatch, predictions=None):\n    \"\"\"This will work with:\n    display_batch_of_images(images)\n    display_batch_of_images(images, predictions)\n    display_batch_of_images((images, labels))\n    display_batch_of_images((images, labels), predictions)\n    \"\"\"\n    # data\n    images, labels = batch_to_numpy_images_and_labels(databatch)\n    if labels is None:\n        labels = [None for _ in enumerate(images)]\n        \n    # auto-squaring: this will drop data that does not fit into square or square-ish rectangle\n    rows = int(math.sqrt(len(images)))\n    cols = len(images)//rows\n        \n    # size and spacing\n    FIGSIZE = 13.0\n    SPACING = 0.1\n    subplot=(rows,cols,1)\n    if rows < cols:\n        plt.figure(figsize=(FIGSIZE,FIGSIZE/cols*rows))\n    else:\n        plt.figure(figsize=(FIGSIZE/rows*cols,FIGSIZE))\n    \n    # display\n    for i, (image, label) in enumerate(zip(images[:rows*cols], labels[:rows*cols])):\n        title = '' if label is None else CLASSES[label]\n        correct = True\n        if predictions is not None:\n            title, correct = title_from_label_and_target(predictions[i], label)\n        dynamic_titlesize = FIGSIZE*SPACING/max(rows,cols)*40+3 # magic formula tested to work from 1x1 to 10x10 images\n        subplot = display_one_flower(image, title, subplot, not correct, titlesize=dynamic_titlesize)\n    \n    #layout\n    plt.tight_layout()\n    if label is None and predictions is None:\n        plt.subplots_adjust(wspace=0, hspace=0)\n    else:\n        plt.subplots_adjust(wspace=SPACING, hspace=SPACING)\n    plt.show()\n\ndef display_confusion_matrix(cmat, score, precision, recall):\n    plt.figure(figsize=(15,15))\n    ax = plt.gca()\n    ax.matshow(cmat, cmap='Reds')\n    ax.set_xticks(range(len(CLASSES)))\n    ax.set_xticklabels(CLASSES, fontdict={'fontsize': 7})\n    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"left\", rotation_mode=\"anchor\")\n    ax.set_yticks(range(len(CLASSES)))\n    ax.set_yticklabels(CLASSES, fontdict={'fontsize': 7})\n    plt.setp(ax.get_yticklabels(), rotation=45, ha=\"right\", rotation_mode=\"anchor\")\n    titlestring = \"\"\n    if score is not None:\n        titlestring += 'f1 = {:.3f} '.format(score)\n    if precision is not None:\n        titlestring += '\\nprecision = {:.3f} '.format(precision)\n    if recall is not None:\n        titlestring += '\\nrecall = {:.3f} '.format(recall)\n    if len(titlestring) > 0:\n        ax.text(101, 1, titlestring, fontdict={'fontsize': 18, 'horizontalalignment':'right', 'verticalalignment':'top', 'color':'#804040'})\n    plt.show()\n    \ndef display_training_curves(training, validation, title, subplot):\n    if subplot%10==1: # set up the subplots on the first call\n        plt.subplots(figsize=(10,10), facecolor='#F0F0F0')\n        plt.tight_layout()\n    ax = plt.subplot(subplot)\n    ax.set_facecolor('#F8F8F8')\n    ax.plot(training)\n    ax.plot(validation)\n    ax.set_title('model '+ title)\n    ax.set_ylabel(title)\n    #ax.set_ylim(0.28,1.05)\n    ax.set_xlabel('epoch')\n    ax.legend(['train', 'valid.'])","metadata":{"id":"t3K6bpscyJau","execution":{"iopub.status.busy":"2021-12-28T12:52:20.466238Z","iopub.execute_input":"2021-12-28T12:52:20.466533Z","iopub.status.idle":"2021-12-28T12:52:20.487805Z","shell.execute_reply.started":"2021-12-28T12:52:20.466504Z","shell.execute_reply":"2021-12-28T12:52:20.486776Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def random_blockout(img, sl=0.1, sh=0.2, rl=0.4):\n    p=random.random()\n    if p>=0.25:\n        h, w, c = img_size, img_size, 3\n        origin_area = tf.cast(h*w, tf.float32)\n\n        e_size_l = tf.cast(tf.round(tf.sqrt(origin_area * sl * rl)), tf.int32)\n        e_size_h = tf.cast(tf.round(tf.sqrt(origin_area * sh / rl)), tf.int32)\n\n        e_height_h = tf.minimum(e_size_h, h)\n        e_width_h = tf.minimum(e_size_h, w)\n\n        erase_height = tf.random.uniform(shape=[], minval=e_size_l, maxval=e_height_h, dtype=tf.int32)\n        erase_width = tf.random.uniform(shape=[], minval=e_size_l, maxval=e_width_h, dtype=tf.int32)\n\n        erase_area = tf.zeros(shape=[erase_height, erase_width, c])\n        erase_area = tf.cast(erase_area, tf.uint8)\n\n        pad_h = h - erase_height\n        pad_top = tf.random.uniform(shape=[], minval=0, maxval=pad_h, dtype=tf.int32)\n        pad_bottom = pad_h - pad_top\n\n        pad_w = w - erase_width\n        pad_left = tf.random.uniform(shape=[], minval=0, maxval=pad_w, dtype=tf.int32)\n        pad_right = pad_w - pad_left\n\n        erase_mask = tf.pad([erase_area], [[0,0],[pad_top, pad_bottom], [pad_left, pad_right], [0,0]], constant_values=1)\n        erase_mask = tf.squeeze(erase_mask, axis=0)\n        erased_img = tf.multiply(tf.cast(img,tf.float32), tf.cast(erase_mask, tf.float32))\n\n        return tf.cast(erased_img, img.dtype)\n    else:\n        return tf.cast(img, img.dtype)","metadata":{"id":"0TK49fvfyMSP","execution":{"iopub.status.busy":"2021-12-28T12:55:55.142308Z","iopub.execute_input":"2021-12-28T12:55:55.143247Z","iopub.status.idle":"2021-12-28T12:55:55.153332Z","shell.execute_reply.started":"2021-12-28T12:55:55.143206Z","shell.execute_reply":"2021-12-28T12:55:55.152679Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def decode_image(image_data):\n    image = tf.image.decode_jpeg(image_data, channels=3)\n    image = tf.cast(image, tf.float32) / 255.0  # convert image to floats in [0, 1] range\n    image = tf.reshape(image, [*IMAGE_SIZE, 3]) # explicit size needed for TPU\n    return image\n\ndef read_labeled_tfrecord(example):\n    LABELED_TFREC_FORMAT = {\n        \"image\": tf.io.FixedLenFeature([], tf.string), # tf.string means bytestring\n        \"class\": tf.io.FixedLenFeature([], tf.int64),  # shape [] means single element\n    }\n    example = tf.io.parse_single_example(example, LABELED_TFREC_FORMAT)\n    image = decode_image(example['image'])\n    label = tf.cast(example['class'], tf.int32)\n    return image, label # returns a dataset of (image, label) pairs\n\ndef read_unlabeled_tfrecord(example):\n    UNLABELED_TFREC_FORMAT = {\n        \"image\": tf.io.FixedLenFeature([], tf.string), # tf.string means bytestring\n        \"id\": tf.io.FixedLenFeature([], tf.string),  # shape [] means single element\n        # class is missing, this competitions's challenge is to predict flower classes for the test dataset\n    }\n    example = tf.io.parse_single_example(example, UNLABELED_TFREC_FORMAT)\n    image = decode_image(example['image'])\n    idnum = example['id']\n    return image, idnum # returns a dataset of image(s)\n\ndef load_dataset(filenames, labeled=True, ordered=False):\n    # Read from TFRecords. For optimal performance, reading from multiple files at once and\n    # disregarding data order. Order does not matter since we will be shuffling the data anyway.\n\n    ignore_order = tf.data.Options()\n    if not ordered:\n        ignore_order.experimental_deterministic = False # disable order, increase speed\n\n    dataset = tf.data.TFRecordDataset(filenames, num_parallel_reads=AUTO) # automatically interleaves reads from multiple files\n    dataset = dataset.with_options(ignore_order) # uses data as soon as it streams in, rather than in its original order\n    dataset = dataset.map(read_labeled_tfrecord if labeled else read_unlabeled_tfrecord, num_parallel_calls=AUTO)\n    # returns a dataset of (image, label) pairs if labeled=True or (image, id) pairs if labeled=False\n    return dataset\n\ndef data_augment(image, label):\n    # data augmentation. Thanks to the dataset.prefetch(AUTO) statement in the next function (below),\n    # this happens essentially for free on TPU. Data pipeline code is executed on the \"CPU\" part\n    # of the TPU while the TPU itself is computing gradients.\n    image = tf.image.random_flip_left_right(image)\n    image= random_blockout(image)\n    image = tf.image.random_saturation(image, 0, 2)\n    return image, label   \n\ndef get_training_dataset():\n    dataset = load_dataset(TRAINING_FILENAMES, labeled=True)\n    #dataset2= load_dataset(TRAINING_FILENAMES, labeled=True)\n    dataset = dataset.map(data_augment, num_parallel_calls=AUTO)\n    #dataset=dataset.concatenate(dataset2)\n    dataset = dataset.repeat() # the training dataset must repeat for several epochs\n    dataset = dataset.shuffle(2048)\n    dataset = dataset.batch(BATCH_SIZE)\n    dataset = dataset.prefetch(AUTO) # prefetch next batch while training (autotune prefetch buffer size)\n    return dataset\n\ndef get_validation_dataset(ordered=False):\n    dataset = load_dataset(VALIDATION_FILENAMES, labeled=True, ordered=ordered)\n    dataset = dataset.batch(BATCH_SIZE)\n    dataset = dataset.cache()\n    dataset = dataset.prefetch(AUTO) # prefetch next batch while training (autotune prefetch buffer size)\n    return dataset\n\ndef get_test_dataset(ordered=False):\n    dataset = load_dataset(TEST_FILENAMES, labeled=False, ordered=ordered)\n    dataset = dataset.batch(BATCH_SIZE)\n    dataset = dataset.prefetch(AUTO) # prefetch next batch while training (autotune prefetch buffer size)\n    return dataset\n\ndef count_data_items(filenames):\n    # the number of data items is written in the name of the .tfrec files, i.e. flowers00-230.tfrec = 230 data items\n    n = [int(re.compile(r\"-([0-9]*)\\.\").search(filename).group(1)) for filename in filenames]\n    return np.sum(n)\n\nNUM_TRAINING_IMAGES = count_data_items(TRAINING_FILENAMES)\nNUM_VALIDATION_IMAGES = count_data_items(VALIDATION_FILENAMES)\nNUM_TEST_IMAGES = count_data_items(TEST_FILENAMES)\nSTEPS_PER_EPOCH = NUM_TRAINING_IMAGES // BATCH_SIZE\nprint('Dataset: {} training images, {} validation images, {} unlabeled test images'.format(NUM_TRAINING_IMAGES, NUM_VALIDATION_IMAGES, NUM_TEST_IMAGES))","metadata":{"id":"OAoAwqLCyPeG","outputId":"0f059f82-67ae-4dcf-dac2-22c2b834c8f0","execution":{"iopub.status.busy":"2021-12-28T12:59:48.0928Z","iopub.execute_input":"2021-12-28T12:59:48.093143Z","iopub.status.idle":"2021-12-28T12:59:48.11101Z","shell.execute_reply.started":"2021-12-28T12:59:48.09311Z","shell.execute_reply":"2021-12-28T12:59:48.110049Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# data dump\nprint(\"Training data shapes:\")\nfor image, label in get_training_dataset().take(3):\n    print(image.numpy().shape, label.numpy().shape)\nprint(\"Training data label examples:\", label.numpy())\nprint(\"Validation data shapes:\")\nfor image, label in get_validation_dataset().take(3):\n    print(image.numpy().shape, label.numpy().shape)\nprint(\"Validation data label examples:\", label.numpy())\nprint(\"Test data shapes:\")\nfor image, idnum in get_test_dataset().take(3):\n    print(image.numpy().shape, idnum.numpy().shape)\nprint(\"Test data IDs:\", idnum.numpy().astype('U')) # U=unicode string","metadata":{"id":"y8x4MnXwySVP","outputId":"5bc32b77-54be-495b-f251-15205ca6e89b","execution":{"iopub.status.busy":"2021-12-28T12:59:56.406841Z","iopub.execute_input":"2021-12-28T12:59:56.40769Z","iopub.status.idle":"2021-12-28T13:00:24.704297Z","shell.execute_reply.started":"2021-12-28T12:59:56.407643Z","shell.execute_reply":"2021-12-28T13:00:24.703562Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Peek at training data\ntraining_dataset = get_training_dataset()\ntraining_dataset = training_dataset.unbatch().batch(20)\ntraining_dataset = training_dataset.shuffle(20)\ntrain_batch = iter(training_dataset)","metadata":{"id":"oPEJQBKYyV1X","execution":{"iopub.status.busy":"2021-12-28T13:06:11.626705Z","iopub.execute_input":"2021-12-28T13:06:11.626973Z","iopub.status.idle":"2021-12-28T13:06:11.719388Z","shell.execute_reply.started":"2021-12-28T13:06:11.626945Z","shell.execute_reply":"2021-12-28T13:06:11.718568Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# run this cell again for next set of images\ndisplay_batch_of_images(next(train_batch))\n","metadata":{"id":"Vu_poIcTyWro","outputId":"9559adf8-d8f6-405f-8609-b6cfab529cf3","execution":{"iopub.status.busy":"2021-12-28T13:04:15.561975Z","iopub.execute_input":"2021-12-28T13:04:15.5626Z","iopub.status.idle":"2021-12-28T13:04:20.435984Z","shell.execute_reply.started":"2021-12-28T13:04:15.562544Z","shell.execute_reply":"2021-12-28T13:04:20.434913Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# peer at test data\ntest_dataset = get_test_dataset()\ntest_dataset = test_dataset.unbatch().batch(20)\ntest_batch = iter(test_dataset)","metadata":{"id":"kPcMMEPeyZbI"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# run this cell again for next set of images\ndisplay_batch_of_images(next(test_batch))","metadata":{"id":"jtwReXH9yaGg","outputId":"41fd5cbb-7056-4120-baca-109eef08fac7"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def lrfn(epoch):\n    LR_START = 0.00001\n    LR_MAX = 0.00005 * strategy.num_replicas_in_sync\n    LR_MIN = 0.00001\n    LR_RAMPUP_EPOCHS = 5\n    LR_SUSTAIN_EPOCHS = 0\n    LR_EXP_DECAY = .8\n    \n    if epoch < LR_RAMPUP_EPOCHS:\n        lr = (LR_MAX - LR_START) / LR_RAMPUP_EPOCHS * epoch + LR_START\n    elif epoch < LR_RAMPUP_EPOCHS + LR_SUSTAIN_EPOCHS:\n        lr = LR_MAX\n    else:\n        lr = (LR_MAX - LR_MIN) * LR_EXP_DECAY**(epoch - LR_RAMPUP_EPOCHS - LR_SUSTAIN_EPOCHS) + LR_MIN\n    return lr\nlr_callback = tf.keras.callbacks.LearningRateScheduler(lrfn, verbose=1)","metadata":{"id":"SjA6Mbc_ybzY"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"!pip install -U efficientnet\nimport efficientnet.tfkeras as efn","metadata":{"id":"iwuVl-okyd6v","outputId":"40f25d71-96b8-488b-88bc-a40399eba4eb"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"with strategy.scope():\n    pretrained_model = tf.keras.applications.Xception(weights='imagenet', include_top=False ,input_shape=[*IMAGE_SIZE, 3])\n    pretrained_model.trainable = True # False = transfer learning, True = fine-tuning\n    \n    model = tf.keras.Sequential([\n        pretrained_model,\n        tf.keras.layers.GlobalAveragePooling2D(),\n        tf.keras.layers.Dense(len(CLASSES), activation='softmax')\n    ])  \n    model.compile(\n    optimizer=tf.keras.optimizers.Adam(),#learning_rate=0.001, beta_1=0.9, beta_2=0.999, amsgrad=False),\n    loss = 'sparse_categorical_crossentropy',\n    metrics=['sparse_categorical_accuracy']\n)\n    model.summary()","metadata":{"id":"ytQLWBkryiHn","outputId":"d0f24d58-5d10-45ed-a356-8dfa0e1db4b3"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"history = model.fit(get_training_dataset(), \n                    steps_per_epoch=STEPS_PER_EPOCH, \n                    epochs=30, \n                    validation_data=get_validation_dataset(), \n                    callbacks = [lr_callback]\n                    #class_weight=class_weights\n                   )","metadata":{"id":"ta7LVHpHyi1_","outputId":"76fc14a3-38ce-490d-cffd-4541a98733f0"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.save(\"xception.h5\")","metadata":{"id":"l_QZ63NkRYKW","outputId":"8339a640-529c-4072-ec9d-98a92ea06b6c"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def display_confusion_matrix(cmat, score, precision, recall):\n    plt.figure(figsize=(15,15))\n    ax = plt.gca()\n    ax.matshow(cmat, cmap='Reds')\n    ax.set_xticks(range(len(CLASSES)))\n    ax.set_xticklabels(CLASSES, fontdict={'fontsize': 7})\n    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"left\", rotation_mode=\"anchor\")\n    ax.set_yticks(range(len(CLASSES)))\n    ax.set_yticklabels(CLASSES, fontdict={'fontsize': 7})\n    plt.setp(ax.get_yticklabels(), rotation=45, ha=\"right\", rotation_mode=\"anchor\")\n    titlestring = \"\"\n    if score is not None:\n        titlestring += 'f1 = {:.3f} '.format(score)\n    if precision is not None:\n        titlestring += '\\nprecision = {:.3f} '.format(precision)\n    if recall is not None:\n        titlestring += '\\nrecall = {:.3f} '.format(recall)\n    if len(titlestring) > 0:\n        ax.text(101, 1, titlestring, fontdict={'fontsize': 18, 'horizontalalignment':'right', 'verticalalignment':'top', 'color':'#804040'})\n    plt.show()","metadata":{"id":"_qNj0pfbpfM5"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cmdataset = get_validation_dataset(ordered=True)\nimages_ds = cmdataset.map(lambda image, label: image)\nlabels_ds = cmdataset.map(lambda image, label: label).unbatch()\n\ncm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy()\ncm_probabilities = model.predict(images_ds)\ncm_predictions = np.argmax(cm_probabilities, axis=-1)\ntemp = []\nfor e in labels_ds:\n    temp.append(e.numpy())\nlabels = range(len(CLASSES))\ncmat = confusion_matrix(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n)\ncmat = (cmat.T / cmat.sum(axis=1)).T # normalize","metadata":{"id":"-HnvVIqVpUiC"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"score = f1_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nprecision = precision_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nrecall = recall_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\ndisplay_confusion_matrix(cmat, score, precision, recall)","metadata":{"id":"de6OJY-vpj_9","outputId":"b104ae3e-ce39-4afc-8e9d-f938598055b1"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model2 = tf.keras.models.load_model(\"/content/drive/MyDrive/Tensorflow Model/densenet201 .h5\")","metadata":{"id":"OjkZ3qiwvICZ"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cmdataset = get_validation_dataset(ordered=True)\nimages_ds = cmdataset.map(lambda image, label: image)\nlabels_ds = cmdataset.map(lambda image, label: label).unbatch()\n\ncm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy()\ncm_probabilities = model2.predict(images_ds)\ncm_predictions = np.argmax(cm_probabilities, axis=-1)\ntemp = []\nfor e in labels_ds:\n    temp.append(e.numpy())\nlabels = range(len(CLASSES))\ncmat = confusion_matrix(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n)\ncmat = (cmat.T / cmat.sum(axis=1)).T # normalize","metadata":{"id":"TV1Ct8xbvgRp"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"score = f1_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nprecision = precision_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nrecall = recall_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\ndisplay_confusion_matrix(cmat, score, precision, recall)","metadata":{"id":"W-BXBEo0vpA6","outputId":"684bfb9c-0c04-4175-a93e-f69b88f35ea6"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model3 = tf.keras.models.load_model(\"/content/drive/MyDrive/Tensorflow Model/efficientnet07.h5\")","metadata":{"id":"NXFWo9m-w5dk"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"cmdataset = get_validation_dataset(ordered=True)\nimages_ds = cmdataset.map(lambda image, label: image)\nlabels_ds = cmdataset.map(lambda image, label: label).unbatch()\n\ncm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy()\ncm_probabilities = model3.predict(images_ds)\ncm_predictions = np.argmax(cm_probabilities, axis=-1)\ntemp = []\nfor e in labels_ds:\n    temp.append(e.numpy())\nlabels = range(len(CLASSES))\ncmat = confusion_matrix(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n)\ncmat = (cmat.T / cmat.sum(axis=1)).T # normalize","metadata":{"id":"AsVjtN0JxDu8"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"score = f1_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nprecision = precision_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\nrecall = recall_score(\n    cm_correct_labels,\n    cm_predictions,\n    labels=labels,\n    average='macro',\n)\ndisplay_confusion_matrix(cmat, score, precision, recall)","metadata":{"id":"R5TWjUUJyZ0M","outputId":"08bbc9ab-2b51-4098-f152-131740e9ca3f"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"m1,m2,m3 = model.predict(images_ds),model2.predict(images_ds),model3.predict(images_ds)","metadata":{"id":"Zwkj3bTq2L1I"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def find_best_alpha(m1,m2):\n    cmdataset = get_validation_dataset(ordered=True) # since we are splitting the dataset and iterating separately on images and labels, order matters.\n    images_ds = cmdataset.map(lambda image, label: image)\n    labels_ds = cmdataset.map(lambda image, label: label).unbatch()\n    cm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy() # get everything as one batch\n    \n    #m1 = model.predict(images_ds)\n    #m2 = model2.predict(images_ds)\n\n    scores = []\n    for alpha in np.linspace(0,1,100):\n        #print(alpha)\n        cm_probabilities = alpha*m1+(1-alpha)*m2\n        cm_predictions = np.argmax(cm_probabilities, axis=-1)\n        scores.append(f1_score(cm_correct_labels, cm_predictions, labels=range(len(CLASSES)), average='macro'))\n\n    print(\"Correct   labels: \", cm_correct_labels.shape, cm_correct_labels)\n    print(\"Predicted labels: \", cm_predictions.shape, cm_predictions)\n    plt.plot(scores)\n\n    best_alpha = np.argmax(scores)/100\n    cm_probabilities = best_alpha*m1+(1-best_alpha)*m2\n    cm_predictions = np.argmax(cm_probabilities, axis=-1)\n    print(\"best_alpha\",best_alpha)\n    return best_alpha","metadata":{"id":"DSnw-8eG2ieO"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def get_cnfmtx(m1,m2):  \n  alpha = find_best_alpha(m1,m2)\n  cmdataset = get_validation_dataset(ordered=True) # since we are splitting the dataset and iterating separately on images and labels, order matters.\n  images_ds = cmdataset.map(lambda image, label: image)\n  labels_ds = cmdataset.map(lambda image, label: label).unbatch()\n  cm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy() # get everything as one batch\n  cm_probabilities = alpha*m1+(1-alpha)*m2\n  cm_predictions = np.argmax(cm_probabilities, axis=-1)\n  labels = range(len(CLASSES))\n  cmat = confusion_matrix(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n  )\n  cmat = (cmat.T / cmat.sum(axis=1)).T # normalize\n  score = f1_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  precision = precision_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  recall = recall_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  display_confusion_matrix(cmat, score, precision, recall)","metadata":{"id":"zF5pFMn29CDX"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"get_cnfmtx(m1,m2)","metadata":{"id":"Y6-iLMS58ccB","outputId":"44445368-208e-4ba4-c822-a62e7ee8b97f"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"get_cnfmtx(m2,m3)","metadata":{"id":"wuf2Mb2u-LCF","outputId":"b54d445b-eb03-4759-eb65-fd005f18d2d1"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"get_cnfmtx(m1,m3)","metadata":{"id":"5mSBQzNN-VC1","outputId":"28d60fb0-7981-48c8-fdd1-67a414594bbd"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def find_best_alpha_three(m1,m2,m3):\n    cmdataset = get_validation_dataset(ordered=True) # since we are splitting the dataset and iterating separately on images and labels, order matters.\n    images_ds = cmdataset.map(lambda image, label: image)\n    labels_ds = cmdataset.map(lambda image, label: label).unbatch()\n    cm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy() # get everything as one batch\n    \n    #m1 = model.predict(images_ds)\n    #m2 = model2.predict(images_ds)\n\n    scores = []\n    for i in range(0,101,10):\n      for j in range(0,101,10):\n        for k in range(0,101,10):\n          if(i+j+k == 100):\n            cm_probabilities = i*m1 + j*m2 + k*m3\n            cm_predictions = np.argmax(cm_probabilities, axis=-1)\n            scores.append(f1_score(cm_correct_labels, cm_predictions, labels=range(len(CLASSES)), average='macro'))\n\n    print(\"Correct   labels: \", cm_correct_labels.shape, cm_correct_labels)\n    print(\"Predicted labels: \", cm_predictions.shape, cm_predictions)\n    plt.plot(scores)\n\n    best_alpha = np.argmax(scores)/100\n    cm_probabilities = best_alpha*m1+(1-best_alpha)*m2\n    cm_predictions = np.argmax(cm_probabilities, axis=-1)\n    print(\"best_alpha\",best_alpha)\n    return best_alpha","metadata":{"id":"2UP9bZ0bKB1c"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from pycm import *\ndef get_cnfmtx_three(m1,m2,m3):  \n  alpha = find_best_alpha_three(m1,m2,m3)*100\n  count,x,y,z = 0,0,0,0\n  for i in range(0,101,10):\n      for j in range(0,101,10):\n        for k in range(0,101,10):\n          if(i+j+k == 100):\n            if(count == alpha):\n              x,y,z = i/100,j/100,k/100\n              #break\n            count +=1\n\n  print(\"w1\",x,\"w2\",y,\"w3\",z)\n  cmdataset = get_validation_dataset(ordered=True) # since we are splitting the dataset and iterating separately on images and labels, order matters.\n  images_ds = cmdataset.map(lambda image, label: image)\n  labels_ds = cmdataset.map(lambda image, label: label).unbatch()\n  cm_correct_labels = next(iter(labels_ds.batch(NUM_VALIDATION_IMAGES))).numpy() # get everything as one batch\n  cm_probabilities = x*m1 + y*m2 + z*m3\n  cm_predictions = np.argmax(cm_probabilities, axis=-1)\n  labels = range(len(CLASSES))\n  cmat = confusion_matrix(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n  )\n  cmat = (cmat.T / cmat.sum(axis=1)).T # normalize\n  \n  \n  cm = ConfusionMatrix(actual_vector=cm_correct_labels, predict_vector=cm_predictions)\n  cm.save_csv(\"confusion_matrix\")\n  \n  score = f1_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  precision = precision_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  recall = recall_score(\n      cm_correct_labels,\n      cm_predictions,\n      labels=labels,\n      average='macro',\n  )\n\n  display_confusion_matrix(cmat, score, precision, recall)","metadata":{"id":"Cf5q__kXMBmH"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"get_cnfmtx_three(m1,m2,m3)","metadata":{"id":"ZWMm6IB-PBWF","outputId":"673d4ac7-932a-4dbd-e275-1c9a6362ab59"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"pip install pycm==3.0","metadata":{"id":"LGRef-XGWGRO","outputId":"08405648-2149-41b9-b117-7733e9dd0eb6"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_ds = get_test_dataset(ordered=True) # since we are splitting the dataset and iterating separately on images and ids, order matters.\n\nprint('Computing predictions...')\ntest_images_ds = test_ds.map(lambda image, idnum: image)\n\nprobabilities = 0.4*model.predict(test_images_ds) + 0.1*model2.predict(test_images_ds) + 0.5*model3.predict(test_images_ds)\n\npredictions = np.argmax(probabilities, axis=-1)\nprint(predictions)\n\nprint('Generating submission.csv file...')\ntest_ids_ds = test_ds.map(lambda image, idnum: idnum).unbatch()\ntest_ids = next(iter(test_ids_ds.batch(NUM_TEST_IMAGES))).numpy().astype('U') # all in one batch\nnp.savetxt('submission.csv', np.rec.fromarrays([test_ids, predictions]), fmt=['%s', '%d'], delimiter=',', header='id,label', comments='')\n!head submission.csv","metadata":{"id":"t205gU6nhSI8","outputId":"519eae14-f313-412d-99b8-d8c7f04b7fa5"},"execution_count":null,"outputs":[]}]}