{"cells":[{"metadata":{},"cell_type":"markdown","source":"> **THIS KERNEL WILL RESULT IN PUBLIC LB AROUND 0.78, BUT IF YOU PLAY AROUND WITH TRANSOFORMATIONS, IMAGE_SIZE, TRAINING CYCLES IT CAN RESULT TO >0.795**"},{"metadata":{},"cell_type":"markdown","source":"This is just a starter kernel, I provide all the basic stuff, you have to figure out how you can utilize it to achieve higher score \nNOTE:\n    - This kernel doesnt use optimized KAPPA\n    - I have trained locally on dataset that was hybrid of new and old competiions (balanced classes), and just uploaded weights\n    - You dont need to do anything fancy to achieve top scores all the hints are in discussions and kernels\n    - I have taken code from public kernels, please let me know if I fogot to mentioned "},{"metadata":{"trusted":true},"cell_type":"code","source":"TTA = False","execution_count":null,"outputs":[]},{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"%reload_ext autoreload\n%autoreload 2\n%matplotlib inline\nimport torch\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nfrom matplotlib import style\nimport seaborn as sns\n%matplotlib inline  \nfrom sklearn.model_selection import StratifiedKFold\nfrom joblib import load, dump\nfrom sklearn.metrics import cohen_kappa_score\nfrom sklearn.metrics import confusion_matrix\nfrom fastai import *\nfrom fastai.vision import *\nfrom fastai.callbacks import *\nfrom torchvision import models as md\nfrom torch import nn\nfrom torch.nn import functional as F\nimport re\nimport math\nimport collections\nfrom functools import partial\nfrom torch.utils import model_zoo\nfrom sklearn import metrics\nfrom collections import Counter\nimport json","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def seed_everything(seed):\n    random.seed(seed)\n    os.environ['PYTHONHASHSEED'] = str(seed)\n    np.random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)\n    torch.backends.cudnn.deterministic = True\n    \nseed_everything(42)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#I could not figure out how to install package in local kernel so i just stole from github =)\n#code stolen from https://github.com/lukemelas/EfficientNet-PyTorch\n\n\n\"\"\"\nThis file contains helper functions for building the model and for loading model parameters.\nThese helper functions are built to mirror those in the official TensorFlow implementation.\n\"\"\"\n\n\n\n# Parameters for the entire model (stem, all blocks, and head)\nGlobalParams = collections.namedtuple('GlobalParams', [\n    'batch_norm_momentum', 'batch_norm_epsilon', 'dropout_rate',\n    'num_classes', 'width_coefficient', 'depth_coefficient',\n    'depth_divisor', 'min_depth', 'drop_connect_rate', 'image_size'])\n\n\n# Parameters for an individual model block\nBlockArgs = collections.namedtuple('BlockArgs', [\n    'kernel_size', 'num_repeat', 'input_filters', 'output_filters',\n    'expand_ratio', 'id_skip', 'stride', 'se_ratio'])\n\n\n# Change namedtuple defaults\nGlobalParams.__new__.__defaults__ = (None,) * len(GlobalParams._fields)\nBlockArgs.__new__.__defaults__ = (None,) * len(BlockArgs._fields)\n\n\ndef relu_fn(x):\n    \"\"\" Swish activation function \"\"\"\n    return x * torch.sigmoid(x)\n\n\ndef round_filters(filters, global_params):\n    \"\"\" Calculate and round number of filters based on depth multiplier. \"\"\"\n    multiplier = global_params.width_coefficient\n    if not multiplier:\n        return filters\n    divisor = global_params.depth_divisor\n    min_depth = global_params.min_depth\n    filters *= multiplier\n    min_depth = min_depth or divisor\n    new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor)\n    if new_filters < 0.9 * filters:  # prevent rounding by more than 10%\n        new_filters += divisor\n    return int(new_filters)\n\n\ndef round_repeats(repeats, global_params):\n    \"\"\" Round number of filters based on depth multiplier. \"\"\"\n    multiplier = global_params.depth_coefficient\n    if not multiplier:\n        return repeats\n    return int(math.ceil(multiplier * repeats))\n\n\ndef drop_connect(inputs, p, training):\n    \"\"\" Drop connect. \"\"\"\n    if not training: return inputs\n    batch_size = inputs.shape[0]\n    keep_prob = 1 - p\n    random_tensor = keep_prob\n    random_tensor += torch.rand([batch_size, 1, 1, 1], dtype=inputs.dtype, device=inputs.device)\n    binary_tensor = torch.floor(random_tensor)\n    output = inputs / keep_prob * binary_tensor\n    return output\n\n\ndef get_same_padding_conv2d(image_size=None):\n    \"\"\" Chooses static padding if you have specified an image size, and dynamic padding otherwise.\n        Static padding is necessary for ONNX exporting of models. \"\"\"\n    if image_size is None:\n        return Conv2dDynamicSamePadding\n    else:\n        return partial(Conv2dStaticSamePadding, image_size=image_size)\n\nclass Conv2dDynamicSamePadding(nn.Conv2d):\n    \"\"\" 2D Convolutions like TensorFlow, for a dynamic image size \"\"\"\n    def __init__(self, in_channels, out_channels, kernel_size, stride=1, dilation=1, groups=1, bias=True):\n        super().__init__(in_channels, out_channels, kernel_size, stride, 0, dilation, groups, bias)\n        self.stride = self.stride if len(self.stride) == 2 else [self.stride[0]]*2\n\n    def forward(self, x):\n        ih, iw = x.size()[-2:]\n        kh, kw = self.weight.size()[-2:]\n        sh, sw = self.stride\n        oh, ow = math.ceil(ih / sh), math.ceil(iw / sw)\n        pad_h = max((oh - 1) * self.stride[0] + (kh - 1) * self.dilation[0] + 1 - ih, 0)\n        pad_w = max((ow - 1) * self.stride[1] + (kw - 1) * self.dilation[1] + 1 - iw, 0)\n        if pad_h > 0 or pad_w > 0:\n            x = F.pad(x, [pad_w//2, pad_w - pad_w//2, pad_h//2, pad_h - pad_h//2])\n        return F.conv2d(x, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups)\n\n\nclass Conv2dStaticSamePadding(nn.Conv2d):\n    \"\"\" 2D Convolutions like TensorFlow, for a fixed image size\"\"\"\n    def __init__(self, in_channels, out_channels, kernel_size, image_size=None, **kwargs):\n        super().__init__(in_channels, out_channels, kernel_size, **kwargs)\n        self.stride = self.stride if len(self.stride) == 2 else [self.stride[0]] * 2\n\n        # Calculate padding based on image size and save it\n        assert image_size is not None\n        ih, iw = image_size if type(image_size) == list else [image_size, image_size]\n        kh, kw = self.weight.size()[-2:]\n        sh, sw = self.stride\n        oh, ow = math.ceil(ih / sh), math.ceil(iw / sw)\n        pad_h = max((oh - 1) * self.stride[0] + (kh - 1) * self.dilation[0] + 1 - ih, 0)\n        pad_w = max((ow - 1) * self.stride[1] + (kw - 1) * self.dilation[1] + 1 - iw, 0)\n        if pad_h > 0 or pad_w > 0:\n            self.static_padding = nn.ZeroPad2d((pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h - pad_h // 2))\n        else:\n            self.static_padding = Identity()\n\n    def forward(self, x):\n        x = self.static_padding(x)\n        x = F.conv2d(x, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups)\n        return x\n\n\nclass Identity(nn.Module):\n    def __init__(self,):\n        super(Identity, self).__init__()\n\n    def forward(self, input):\n        return input\n\n\n########################################################################\n############## HELPERS FUNCTIONS FOR LOADING MODEL PARAMS ##############\n########################################################################\n\n\ndef efficientnet_params(model_name):\n    \"\"\" Map EfficientNet model name to parameter coefficients. \"\"\"\n    params_dict = {\n        # Coefficients:   width,depth,res,dropout\n        'efficientnet-b0': (1.0, 1.0, 224, 0.2),\n        'efficientnet-b1': (1.0, 1.1, 240, 0.2),\n        'efficientnet-b2': (1.1, 1.2, 260, 0.3),\n        'efficientnet-b3': (1.2, 1.4, 300, 0.3),\n        'efficientnet-b4': (1.4, 1.8, 380, 0.4),\n        'efficientnet-b5': (1.6, 2.2, 456, 0.4),\n        'efficientnet-b6': (1.8, 2.6, 528, 0.5),\n        'efficientnet-b7': (2.0, 3.1, 600, 0.5),\n    }\n    return params_dict[model_name]\n\n\nclass BlockDecoder(object):\n    \"\"\" Block Decoder for readability, straight from the official TensorFlow repository \"\"\"\n\n    @staticmethod\n    def _decode_block_string(block_string):\n        \"\"\" Gets a block through a string notation of arguments. \"\"\"\n        assert isinstance(block_string, str)\n\n        ops = block_string.split('_')\n        options = {}\n        for op in ops:\n            splits = re.split(r'(\\d.*)', op)\n            if len(splits) >= 2:\n                key, value = splits[:2]\n                options[key] = value\n\n        # Check stride\n        assert (('s' in options and len(options['s']) == 1) or\n                (len(options['s']) == 2 and options['s'][0] == options['s'][1]))\n\n        return BlockArgs(\n            kernel_size=int(options['k']),\n            num_repeat=int(options['r']),\n            input_filters=int(options['i']),\n            output_filters=int(options['o']),\n            expand_ratio=int(options['e']),\n            id_skip=('noskip' not in block_string),\n            se_ratio=float(options['se']) if 'se' in options else None,\n            stride=[int(options['s'][0])])\n\n    @staticmethod\n    def _encode_block_string(block):\n        \"\"\"Encodes a block to a string.\"\"\"\n        args = [\n            'r%d' % block.num_repeat,\n            'k%d' % block.kernel_size,\n            's%d%d' % (block.strides[0], block.strides[1]),\n            'e%s' % block.expand_ratio,\n            'i%d' % block.input_filters,\n            'o%d' % block.output_filters\n        ]\n        if 0 < block.se_ratio <= 1:\n            args.append('se%s' % block.se_ratio)\n        if block.id_skip is False:\n            args.append('noskip')\n        return '_'.join(args)\n\n    @staticmethod\n    def decode(string_list):\n        \"\"\"\n        Decodes a list of string notations to specify blocks inside the network.\n\n        :param string_list: a list of strings, each string is a notation of block\n        :return: a list of BlockArgs namedtuples of block args\n        \"\"\"\n        assert isinstance(string_list, list)\n        blocks_args = []\n        for block_string in string_list:\n            blocks_args.append(BlockDecoder._decode_block_string(block_string))\n        return blocks_args\n\n    @staticmethod\n    def encode(blocks_args):\n        \"\"\"\n        Encodes a list of BlockArgs to a list of strings.\n\n        :param blocks_args: a list of BlockArgs namedtuples of block args\n        :return: a list of strings, each string is a notation of block\n        \"\"\"\n        block_strings = []\n        for block in blocks_args:\n            block_strings.append(BlockDecoder._encode_block_string(block))\n        return block_strings\n\n\ndef efficientnet(width_coefficient=None, depth_coefficient=None, dropout_rate=0.2,\n                 drop_connect_rate=0.2, image_size=None, num_classes=1000):\n    \"\"\" Creates a efficientnet model. \"\"\"\n\n    blocks_args = [\n        'r1_k3_s11_e1_i32_o16_se0.25', 'r2_k3_s22_e6_i16_o24_se0.25',\n        'r2_k5_s22_e6_i24_o40_se0.25', 'r3_k3_s22_e6_i40_o80_se0.25',\n        'r3_k5_s11_e6_i80_o112_se0.25', 'r4_k5_s22_e6_i112_o192_se0.25',\n        'r1_k3_s11_e6_i192_o320_se0.25',\n    ]\n    blocks_args = BlockDecoder.decode(blocks_args)\n\n    global_params = GlobalParams(\n        batch_norm_momentum=0.99,\n        batch_norm_epsilon=1e-3,\n        dropout_rate=dropout_rate,\n        drop_connect_rate=drop_connect_rate,\n        # data_format='channels_last',  # removed, this is always true in PyTorch\n        num_classes=num_classes,\n        width_coefficient=width_coefficient,\n        depth_coefficient=depth_coefficient,\n        depth_divisor=8,\n        min_depth=None,\n        image_size=image_size,\n    )\n\n    return blocks_args, global_params\n\n\ndef get_model_params(model_name, override_params):\n    \"\"\" Get the block args and global params for a given model \"\"\"\n    if model_name.startswith('efficientnet'):\n        w, d, s, p = efficientnet_params(model_name)\n        # note: all models have drop connect rate = 0.2\n        blocks_args, global_params = efficientnet(\n            width_coefficient=w, depth_coefficient=d, dropout_rate=p, image_size=s)\n    else:\n        raise NotImplementedError('model name is not pre-defined: %s' % model_name)\n    if override_params:\n        # ValueError will be raised here if override_params has fields not included in global_params.\n        global_params = global_params._replace(**override_params)\n    return blocks_args, global_params\n\n\nurl_map = {\n    'efficientnet-b0': 'http://storage.googleapis.com/public-models/efficientnet-b0-08094119.pth',\n    'efficientnet-b1': 'http://storage.googleapis.com/public-models/efficientnet-b1-dbc7070a.pth',\n    'efficientnet-b2': 'http://storage.googleapis.com/public-models/efficientnet-b2-27687264.pth',\n    'efficientnet-b3': 'http://storage.googleapis.com/public-models/efficientnet-b3-c8376fa2.pth',\n    'efficientnet-b4': 'http://storage.googleapis.com/public-models/efficientnet-b4-e116e8b3.pth',\n    'efficientnet-b5': 'http://storage.googleapis.com/public-models/efficientnet-b5-586e6cc6.pth',\n}\n\ndef load_pretrained_weights(model, model_name, load_fc=True):\n    \"\"\" Loads pretrained weights, and downloads if loading for the first time. \"\"\"\n    state_dict = model_zoo.load_url(url_map[model_name])\n    if load_fc:\n        model.load_state_dict(state_dict)\n    else:\n        state_dict.pop('_fc.weight')\n        state_dict.pop('_fc.bias')\n        res = model.load_state_dict(state_dict, strict=False)\n        assert str(res.missing_keys) == str(['_fc.weight', '_fc.bias']), 'issue loading pretrained weights'\n    print('Loaded pretrained weights for {}'.format(model_name))\n    \n    \nclass MBConvBlock(nn.Module):\n    \"\"\"\n    Mobile Inverted Residual Bottleneck Block\n\n    Args:\n        block_args (namedtuple): BlockArgs, see above\n        global_params (namedtuple): GlobalParam, see above\n\n    Attributes:\n        has_se (bool): Whether the block contains a Squeeze and Excitation layer.\n    \"\"\"\n\n    def __init__(self, block_args, global_params):\n        super().__init__()\n        self._block_args = block_args\n        self._bn_mom = 1 - global_params.batch_norm_momentum\n        self._bn_eps = global_params.batch_norm_epsilon\n        self.has_se = (self._block_args.se_ratio is not None) and (0 < self._block_args.se_ratio <= 1)\n        self.id_skip = block_args.id_skip  # skip connection and drop connect\n\n        # Get static or dynamic convolution depending on image size\n        Conv2d = get_same_padding_conv2d(image_size=global_params.image_size)\n\n        # Expansion phase\n        inp = self._block_args.input_filters  # number of input channels\n        oup = self._block_args.input_filters * self._block_args.expand_ratio  # number of output channels\n        if self._block_args.expand_ratio != 1:\n            self._expand_conv = Conv2d(in_channels=inp, out_channels=oup, kernel_size=1, bias=False)\n            self._bn0 = nn.BatchNorm2d(num_features=oup, momentum=self._bn_mom, eps=self._bn_eps)\n\n        # Depthwise convolution phase\n        k = self._block_args.kernel_size\n        s = self._block_args.stride\n        self._depthwise_conv = Conv2d(\n            in_channels=oup, out_channels=oup, groups=oup,  # groups makes it depthwise\n            kernel_size=k, stride=s, bias=False)\n        self._bn1 = nn.BatchNorm2d(num_features=oup, momentum=self._bn_mom, eps=self._bn_eps)\n\n        # Squeeze and Excitation layer, if desired\n        if self.has_se:\n            num_squeezed_channels = max(1, int(self._block_args.input_filters * self._block_args.se_ratio))\n            self._se_reduce = Conv2d(in_channels=oup, out_channels=num_squeezed_channels, kernel_size=1)\n            self._se_expand = Conv2d(in_channels=num_squeezed_channels, out_channels=oup, kernel_size=1)\n\n        # Output phase\n        final_oup = self._block_args.output_filters\n        self._project_conv = Conv2d(in_channels=oup, out_channels=final_oup, kernel_size=1, bias=False)\n        self._bn2 = nn.BatchNorm2d(num_features=final_oup, momentum=self._bn_mom, eps=self._bn_eps)\n\n    def forward(self, inputs, drop_connect_rate=None):\n        \"\"\"\n        :param inputs: input tensor\n        :param drop_connect_rate: drop connect rate (float, between 0 and 1)\n        :return: output of block\n        \"\"\"\n\n        # Expansion and Depthwise Convolution\n        x = inputs\n        if self._block_args.expand_ratio != 1:\n            x = relu_fn(self._bn0(self._expand_conv(inputs)))\n        x = relu_fn(self._bn1(self._depthwise_conv(x)))\n\n        # Squeeze and Excitation\n        if self.has_se:\n            x_squeezed = F.adaptive_avg_pool2d(x, 1)\n            x_squeezed = self._se_expand(relu_fn(self._se_reduce(x_squeezed)))\n            x = torch.sigmoid(x_squeezed) * x\n\n        x = self._bn2(self._project_conv(x))\n\n        # Skip connection and drop connect\n        input_filters, output_filters = self._block_args.input_filters, self._block_args.output_filters\n        if self.id_skip and self._block_args.stride == 1 and input_filters == output_filters:\n            if drop_connect_rate:\n                x = drop_connect(x, p=drop_connect_rate, training=self.training)\n            x = x + inputs  # skip connection\n        return x\n\n\nclass EfficientNet(nn.Module):\n    \"\"\"\n    An EfficientNet model. Most easily loaded with the .from_name or .from_pretrained methods\n\n    Args:\n        blocks_args (list): A list of BlockArgs to construct blocks\n        global_params (namedtuple): A set of GlobalParams shared between blocks\n\n    Example:\n        model = EfficientNet.from_pretrained('efficientnet-b0')\n\n    \"\"\"\n\n    def __init__(self, blocks_args=None, global_params=None):\n        super().__init__()\n        assert isinstance(blocks_args, list), 'blocks_args should be a list'\n        assert len(blocks_args) > 0, 'block args must be greater than 0'\n        self._global_params = global_params\n        self._blocks_args = blocks_args\n\n        # Get static or dynamic convolution depending on image size\n        Conv2d = get_same_padding_conv2d(image_size=global_params.image_size)\n\n        # Batch norm parameters\n        bn_mom = 1 - self._global_params.batch_norm_momentum\n        bn_eps = self._global_params.batch_norm_epsilon\n\n        # Stem\n        in_channels = 3  # rgb\n        out_channels = round_filters(32, self._global_params)  # number of output channels\n        self._conv_stem = Conv2d(in_channels, out_channels, kernel_size=3, stride=2, bias=False)\n        self._bn0 = nn.BatchNorm2d(num_features=out_channels, momentum=bn_mom, eps=bn_eps)\n\n        # Build blocks\n        self._blocks = nn.ModuleList([])\n        for block_args in self._blocks_args:\n\n            # Update block input and output filters based on depth multiplier.\n            block_args = block_args._replace(\n                input_filters=round_filters(block_args.input_filters, self._global_params),\n                output_filters=round_filters(block_args.output_filters, self._global_params),\n                num_repeat=round_repeats(block_args.num_repeat, self._global_params)\n            )\n\n            # The first block needs to take care of stride and filter size increase.\n            self._blocks.append(MBConvBlock(block_args, self._global_params))\n            if block_args.num_repeat > 1:\n                block_args = block_args._replace(input_filters=block_args.output_filters, stride=1)\n            for _ in range(block_args.num_repeat - 1):\n                self._blocks.append(MBConvBlock(block_args, self._global_params))\n\n        # Head\n        in_channels = block_args.output_filters  # output of final block\n        out_channels = round_filters(1280, self._global_params)\n        self._conv_head = Conv2d(in_channels, out_channels, kernel_size=1, bias=False)\n        self._bn1 = nn.BatchNorm2d(num_features=out_channels, momentum=bn_mom, eps=bn_eps)\n\n        # Final linear layer\n        self._dropout = self._global_params.dropout_rate\n        self._fc = nn.Linear(out_channels, self._global_params.num_classes)\n\n    def extract_features(self, inputs):\n        \"\"\" Returns output of the final convolution layer \"\"\"\n\n        # Stem\n        x = relu_fn(self._bn0(self._conv_stem(inputs)))\n\n        # Blocks\n        for idx, block in enumerate(self._blocks):\n            drop_connect_rate = self._global_params.drop_connect_rate\n            if drop_connect_rate:\n                drop_connect_rate *= float(idx) / len(self._blocks)\n            x = block(x, drop_connect_rate=drop_connect_rate)\n\n        # Head\n        x = relu_fn(self._bn1(self._conv_head(x)))\n\n        return x\n\n    def forward(self, inputs):\n        \"\"\" Calls extract_features to extract features, applies final linear layer, and returns logits. \"\"\"\n\n        # Convolution layers\n        x = self.extract_features(inputs)\n\n        # Pooling and final linear layer\n        x = F.adaptive_avg_pool2d(x, 1).squeeze(-1).squeeze(-1)\n        if self._dropout:\n            x = F.dropout(x, p=self._dropout, training=self.training)\n        x = self._fc(x)\n        return x\n\n    @classmethod\n    def from_name(cls, model_name, override_params=None):\n        cls._check_model_name_is_valid(model_name)\n        blocks_args, global_params = get_model_params(model_name, override_params)\n        return EfficientNet(blocks_args, global_params)\n\n    @classmethod\n    def from_pretrained(cls, model_name, num_classes=1000):\n        model = EfficientNet.from_name(model_name, override_params={'num_classes': num_classes})\n        return model\n\n    @classmethod\n    def get_image_size(cls, model_name):\n        cls._check_model_name_is_valid(model_name)\n        _, _, res, _ = efficientnet_params(model_name)\n        return res\n\n    @classmethod\n    def _check_model_name_is_valid(cls, model_name, also_need_pretrained_weights=False):\n        \"\"\" Validates model name. None that pretrained weights are only available for\n        the first four models (efficientnet-b{i} for i in 0,1,2,3) at the moment. \"\"\"\n        num_models = 4 if also_need_pretrained_weights else 8\n        valid_models = ['efficientnet_b'+str(i) for i in range(num_models)]\n        if model_name.replace('-','_') not in valid_models:\n            raise ValueError('model_name should be one of: ' + ', '.join(valid_models))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_df():\n    base_image_dir = os.path.join('..', 'input/aptos2019-blindness-detection/')\n    train_dir = os.path.join(base_image_dir,'train_images/')\n    df = pd.read_csv(os.path.join(base_image_dir, 'train.csv'))\n    df['path'] = df['id_code'].map(lambda x: os.path.join(train_dir,'{}.png'.format(x)))\n    df = df.drop(columns=['id_code'])\n    df = df.sample(frac=1).reset_index(drop=True) #shuffle dataframe\n    test_df = pd.read_csv('../input/aptos2019-blindness-detection/sample_submission.csv')\n    return df, test_df\n\ndf, test_df = get_df()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def qk(y_pred, y):\n    return torch.tensor(cohen_kappa_score(torch.round(y_pred), y, weights='quadratic'), device='cuda:0')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Original: https://www.kaggle.com/abhishek/optimizer-for-quadratic-weighted-kappa\n# Improved: https://www.kaggle.com/naveenasaithambi/optimizedrounder-improved\n# Rounds the predictions. Rounding coefficients can be optimised.\n# (e.g. 2.31 could round either up or down depending on the coefficients)\n\nimport scipy as sp\n\nclass OptimizedRounder(object):\n    def __init__(self):\n        self.coef_ = 0\n\n    def _kappa_loss(self, coef, X, y):\n        X_p = np.copy(X)\n        for i, pred in enumerate(X_p):\n            if pred < coef[0]:\n                X_p[i] = 0\n            elif pred >= coef[0] and pred < coef[1]:\n                X_p[i] = 1\n            elif pred >= coef[1] and pred < coef[2]:\n                X_p[i] = 2\n            elif pred >= coef[2] and pred < coef[3]:\n                X_p[i] = 3\n            else:\n                X_p[i] = 4\n\n        ll = metrics.cohen_kappa_score(y, X_p, weights='quadratic')\n        return -ll\n\n    def fit(self, X, y):\n        loss_partial = partial(self._kappa_loss, X=X, y=y)\n        initial_coef = [0.5, 1.5, 2.5, 3.5]\n        self.coef_ = sp.optimize.minimize(loss_partial, initial_coef, method='nelder-mead')\n        print(-loss_partial(self.coef_['x']))\n\n    def predict(self, X, coef):\n        X_p = np.copy(X)\n        for i, pred in enumerate(X_p):\n            if pred < coef[0]:\n                X_p[i] = 0\n            elif pred >= coef[0] and pred < coef[1]:\n                X_p[i] = 1\n            elif pred >= coef[1] and pred < coef[2]:\n                X_p[i] = 2\n            elif pred >= coef[2] and pred < coef[3]:\n                X_p[i] = 3\n            else:\n                X_p[i] = 4\n        return X_p\n\n    def coefficients(self):\n        return self.coef_['x']\n   \n# To optimize adapt the code below:\n# rounder = OptimizedRounder()\n# valid_predictions = [0.9, 0.4, 0.3, 1.2, 1.6, 0.4, 2.3, 2.7, 3.3, 3.4, 3.7, 3.9, 4.5, 4.2, 4.3]\n# targets = [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4]\n# rounder.fit(np.repeat(valid_predictions), np.repeat(targets))\n# coefficients = rounder.coefficients()\n# print(coefficients)\n# valid_predictions = rounder.predict(valid_predictions, coefficients)\n# print(valid_predictions)\n#test_predictions = rounder.predict(test_predictions, coefficients)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# TTA (Test-time augmentation)\nfrom fastai.core import *\nfrom fastai.basic_data import *\nfrom fastai.basic_train import *\nfrom fastai.torch_core import *\ndef _tta_only(learn:Learner, ds_type:DatasetType=DatasetType.Valid, num_pred:int=10) -> Iterator[List[Tensor]]:\n    \"Computes the outputs for several augmented inputs for TTA\"\n    dl = learn.dl(ds_type)\n    ds = dl.dataset\n    old = ds.tfms\n    aug_tfms = [o for o in learn.data.train_ds.tfms]\n    try:\n        pbar = master_bar(range(num_pred))\n        for i in pbar:\n            ds.tfms = aug_tfms\n            yield get_preds(learn.model, dl, pbar=pbar)[0]\n    finally: ds.tfms = old\n\nLearner.tta_only = _tta_only\n\ndef _TTA(learn:Learner, beta:float=0, ds_type:DatasetType=DatasetType.Valid, num_pred:int=10, with_loss:bool=False) -> Tensors:\n    \"Applies TTA to predict on `ds_type` dataset.\"\n    preds,y = learn.get_preds(ds_type)\n    all_preds = list(learn.tta_only(ds_type=ds_type, num_pred=num_pred))\n    avg_preds = torch.stack(all_preds).mean(0)\n    if beta is None: return preds,avg_preds,y\n    else:            \n        final_preds = preds*beta + avg_preds*(1-beta)\n        if with_loss: \n            with NoneReduceOnCPU(learn.loss_func) as lf: loss = lf(final_preds, y)\n            return final_preds, y, loss\n        return final_preds, y\n\nLearner.TTA = _TTA","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def run_subm(learn, rounder_coefficients=[0.5, 1.5, 2.5, 3.5], tta=True):\n    rounder = OptimizedRounder()\n    if tta:\n        preds,y = learn.TTA(ds_type=DatasetType.Test)\n    else:\n        preds, y = learn.get_preds(DatasetType.Test)\n    tst_pred = rounder.predict(preds, rounder_coefficients)\n    test_df.diagnosis = tst_pred.astype(int)\n    test_df.to_csv('submission.csv',index=False)\n    print('done')\n    \n    return test_df","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"####################\n# Data preparation #\n####################\n\n# TODO: Play around with tfms and image sizes\nbs = 64\nsz = 224\ntfms = get_transforms(do_flip=True,flip_vert=True,max_rotate=360,max_warp=0,max_zoom=1.1,max_lighting=0.1,p_lighting=0.5)\n\n# Preparing data\ndata = (ImageList.from_df(df=df,path='./',cols='path') \n        .split_by_rand_pct(0.2) \n        .label_from_df(cols='diagnosis',label_cls=FloatList) \n        .transform(tfms,size=sz,resize_method=ResizeMethod.SQUISH,padding_mode='zeros') \n        .databunch(bs=bs,num_workers=4) \n        .normalize(imagenet_stats)  \n       )\n\ndata.show_batch(rows=3, figsize=(7,6), ds_type=DatasetType.Train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Making model\nmd_ef = EfficientNet.from_pretrained('efficientnet-b5', num_classes=1)\n\n# Preparing model with train data\nlearn = Learner(data, \n                md_ef, \n                metrics = [qk], \n                callback_fns=[\n                    BnFreeze,\n#                   StratifiedBatchCallback,\n#                   partial(GradientClipping, clip=0.2),\n                    partial(SaveModelCallback, monitor='quad_kappa', name='bestmodel')\n                ],\n                model_dir=\"models\",\n               )\nif TTA:\n    learn = learn.to_fp32() # runs out of mem, therefore no TTA possible\nelse:\n    learn = learn.to_fp16() \n\n\n# Giving test data to model\nlearn.data.add_test(ImageList.from_df(test_df,\n                                      '../input/aptos2019-blindness-detection',\n                                      folder='test_images',\n                                      suffix='.png'))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#######################\n# Loading pre-trained #\n#######################\n#copying weighst to the local directory \n!mkdir models\n!cp '../input/kaggle-public/abcdef.pth' 'models'\n\nlearn.load('abcdef'); # Initial model taken from fork","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# learn.lr_find(start_lr=1e-5, end_lr=1e1, wd=5e-3)\n# learn.recorder.plot(suggestion=True)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"##############\n# Finetuning #\n##############\nlearn.fit_one_cycle(5, 1e-4, div_factor=20)\nlearn.recorder.plot_losses()\nlearn.recorder.plot_metrics()\nlearn.export('capggle2')\nlearn.save('cappgle2-save')\n# learn.load('../input/models/cappgle2-save'); # Justinas' trained model\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Load best model found during fine-tunning\nlearn.load('bestmodel')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# # Getting valid predictions from the validation dataset\n# valid_preds = learn.get_preds(ds_type=DatasetType.Valid)\n# _ = pd.DataFrame(valid_preds[0].numpy().flatten()).hist()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Getting valid predictions from the validation dataset\nvalid_preds = learn.TTA(ds_type=DatasetType.Valid)\n_ = pd.DataFrame(valid_preds[0].numpy().flatten()).hist()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Optimise rounder\nrounder = OptimizedRounder()\nrounder.fit(valid_preds[0], valid_preds[1])\nrounder_coefficients = rounder.coefficients()\nprint(rounder_coefficients)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#############\n# Inference #\n#############\ntest_df_tta = run_subm(learn=learn, rounder_coefficients=rounder_coefficients, tta=TTA) # No TTA, memory issues","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# test_df['diagnosis'].value_counts()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_df_tta['diagnosis'].value_counts() # TTA","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.4","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}