Image_classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "agreed-finish"
   },
   "outputs": [],
   "source": [
    "# Cactus classification ... this developed using the cactus dataset \n",
    "# used for a binary classifier. The dataset is taken from: Kaggle: cactus images:\n",
    "# https://www.kaggle.com/c/aerial-cactus-identification/data\n",
    "# Download the zip file from the link and create a directory \"dataFromKaggle\" in the directory\n",
    "# Where this .jpynb file is run from and unzip the dataset of cactus images.\n",
    "# https://towardsdatascience.com/machine-learning-image-classification-kaggle-competition-aerial-cactus-identification-3f720aa9e51c\n",
    "\n",
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import os\n",
    "import math\n",
    "from os import listdir\n",
    "\n",
    "# Matplotlib is a graphing library\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.image as mpimg\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from tensorflow.keras import layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "brave-graduate",
    "outputId": "795e15d2-a1ca-44da-bf6a-74d9f914e6b2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total entries: 17500\n",
      "                                      has_cactus\n",
      "id                                              \n",
      "0004be2cfeaba1c0361d39e2b000257b.jpg           1\n",
      "000c8a36845c0208e833c79c1bffedd1.jpg           1\n",
      "000d1e9a533f62e55c289303b072733d.jpg           1\n",
      "0011485b40695e9138e92d0b3fb55128.jpg           1\n",
      "0014d7a11e90b62848904c1418fc8cf2.jpg           1\n",
      "0017c3c18ddd57a2ea6f9848c79d83d2.jpg           1\n",
      "002134abf28af54575c18741b89dd2a4.jpg           0\n",
      "0024320f43bdd490562246435af4f90b.jpg           0\n",
      "002930423b9840e67e5a54afd4768a1e.jpg           1\n",
      "00351838ebf6dff6e53056e00a1e307c.jpg           1\n"
     ]
    }
   ],
   "source": [
    "# https://www.kaggle.com/vladminzatu/cactus-detection-with-tensorflow-2-0\n",
    "train_lables_df = pd.read_csv('dataFromKaggle/aerial-cactus-identification/train.csv',\\\n",
    "                              index_col='id')\n",
    "print('Total entries: ' + str(train_lables_df.size))\n",
    "print(train_lables_df.head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "maritime-contrast",
    "outputId": "38792183-e777-4cd0-9706-78a567343d49"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    13136\n",
       "0     4364\n",
       "Name: has_cactus, dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_lables_df['has_cactus'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "legal-transition",
    "outputId": "20eec5b2-567c-4a1f-b51e-46cfc2225562"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17500"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dir = 'dataFromKaggle/aerial-cactus-identification/train/train/'\n",
    "train_imgs = os.listdir(train_dir)\n",
    "nr_train_images = len(train_imgs)\n",
    "nr_train_images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "restricted-boundary"
   },
   "outputs": [],
   "source": [
    "def get_train_image_path(id):\n",
    "    return train_dir + id\n",
    "\n",
    "def draw_cactus_image(id, ax):\n",
    "    path = get_train_image_path(id)\n",
    "    img = mpimg.imread(path)\n",
    "    plt.imshow(img)\n",
    "    ax.set_title('Label: ' + str(train_lables_df.loc[id]['has_cactus']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "square-reynolds",
    "outputId": "156d60d4-f803-4e7b-a73b-b7479dd9cb01"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0004be2cfeaba1c0361d39e2b000257b.jpg\n",
      "000c8a36845c0208e833c79c1bffedd1.jpg\n",
      "000d1e9a533f62e55c289303b072733d.jpg\n",
      "0011485b40695e9138e92d0b3fb55128.jpg\n",
      "0014d7a11e90b62848904c1418fc8cf2.jpg\n",
      "0017c3c18ddd57a2ea6f9848c79d83d2.jpg\n",
      "002134abf28af54575c18741b89dd2a4.jpg\n",
      "0024320f43bdd490562246435af4f90b.jpg\n",
      "002930423b9840e67e5a54afd4768a1e.jpg\n",
      "00351838ebf6dff6e53056e00a1e307c.jpg\n",
      "003519dd841a97ed16481fa0657df04d.jpg\n",
      "003bb64852016d9c87871ddd8e25ab03.jpg\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(20,20))\n",
    "for i in range(12):\n",
    "    ax = fig.add_subplot(3, 4, i + 1)\n",
    "    print(train_imgs[i])\n",
    "    draw_cactus_image(train_imgs[i], ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "personalized-climb",
    "outputId": "b7a153b6-99f4-448c-f4cf-dec5fe31ef46"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'list'>\n",
      "17500\n",
      "Original: image path and one label\n",
      "dataFromKaggle/aerial-cactus-identification/train/0004be2cfeaba1c0361d39e2b000257b.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/000c8a36845c0208e833c79c1bffedd1.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/000d1e9a533f62e55c289303b072733d.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/0011485b40695e9138e92d0b3fb55128.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/0014d7a11e90b62848904c1418fc8cf2.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/0017c3c18ddd57a2ea6f9848c79d83d2.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/002134abf28af54575c18741b89dd2a4.jpg 0\n",
      "dataFromKaggle/aerial-cactus-identification/train/0024320f43bdd490562246435af4f90b.jpg 0\n",
      "dataFromKaggle/aerial-cactus-identification/train/002930423b9840e67e5a54afd4768a1e.jpg 1\n",
      "dataFromKaggle/aerial-cactus-identification/train/00351838ebf6dff6e53056e00a1e307c.jpg 1\n",
      "Converted: image path and two labels\n",
      "dataFromKaggle/aerial-cactus-identification/train/0004be2cfeaba1c0361d39e2b000257b.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/000c8a36845c0208e833c79c1bffedd1.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/000d1e9a533f62e55c289303b072733d.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/0011485b40695e9138e92d0b3fb55128.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/0014d7a11e90b62848904c1418fc8cf2.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/0017c3c18ddd57a2ea6f9848c79d83d2.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/002134abf28af54575c18741b89dd2a4.jpg [1, 0]\n",
      "dataFromKaggle/aerial-cactus-identification/train/0024320f43bdd490562246435af4f90b.jpg [1, 0]\n",
      "dataFromKaggle/aerial-cactus-identification/train/002930423b9840e67e5a54afd4768a1e.jpg [0, 1]\n",
      "dataFromKaggle/aerial-cactus-identification/train/00351838ebf6dff6e53056e00a1e307c.jpg [0, 1]\n"
     ]
    }
   ],
   "source": [
    "train_image_paths = [train_dir + ti for ti in train_imgs]\n",
    "train_image_labels = [train_lables_df.loc[ti]['has_cactus'] for ti in train_imgs]\n",
    "\n",
    "print(type(train_image_labels))\n",
    "print(len(train_image_labels))\n",
    "\n",
    "train_image_two_labels = []\n",
    "\n",
    "# Since we want to use multi-label classifier instead of binary classifier\n",
    "# The binary classification labels are converted into two label classifiers\n",
    "# The labels are \"not a cactus\" and \"a cactus\"\n",
    "for i in range(len(train_image_labels)):\n",
    "    if (train_image_labels[i]):\n",
    "        train_image_two_labels.append([0, 1]) # 2nd label, a Cactus\n",
    "    else:\n",
    "        train_image_two_labels.append([1, 0]) # 1st label not a cactus      \n",
    "\n",
    "print('Original: image path and one label')\n",
    "for i in range(10):\n",
    "    print(train_image_paths[i], train_image_labels[i])\n",
    "\n",
    "print('Converted: image path and two labels')\n",
    "for i in range(10):\n",
    "    print(train_image_paths[i], train_image_two_labels[i])   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "dominican-perception"
   },
   "outputs": [],
   "source": [
    "def img_to_tensor(img_path):\n",
    "    img_tensor = tf.cast(tf.image.decode_image(tf.io.read_file(img_path)), tf.float32)\n",
    "    img_tensor /= 255.0 # normalized to [0.0 to 1.0]\n",
    "    return img_tensor\n",
    "\n",
    "img_to_tensor(train_image_paths[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "stylish-member"
   },
   "outputs": [],
   "source": [
    "X_train, X_valid, y_train, y_valid = train_test_split(train_image_paths,\n",
    "                                                      train_image_two_labels,\n",
    "                                                      test_size=0.2)\n",
    "\n",
    "def process_image_in_record(path, label):\n",
    "    return img_to_tensor(path), label\n",
    "\n",
    "# tf.data.dataset manual:\n",
    "# https://www.tensorflow.org/api_docs/python/tf/data/Dataset\n",
    "def build_training_dataset(paths, labels, batch_size=32):\n",
    "    ds = tf.data.Dataset.from_tensor_slices((paths, labels))\n",
    "    ds = ds.map(process_image_in_record)\n",
    "    ds = ds.shuffle(buffer_size=len(paths))\n",
    "    ds = ds.repeat()\n",
    "    ds = ds.batch(batch_size)\n",
    "    ds = ds.prefetch(tf.data.experimental.AUTOTUNE)\n",
    "    return ds\n",
    "\n",
    "def build_validation_dataset(paths, labels, batch_size=32):\n",
    "    ds = tf.data.Dataset.from_tensor_slices((paths, labels))\n",
    "    ds = ds.map(process_image_in_record)\n",
    "    ds = ds.batch(batch_size)\n",
    "    return ds\n",
    "\n",
    "train_ds = build_training_dataset(X_train, y_train)\n",
    "validation_ds = build_validation_dataset(X_valid, y_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "professional-breeding",
    "outputId": "f18b5276-1214-4131-ac36-67a7d1c91e9e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor(\n",
      "[[[[0.38431373 0.31764707 0.34117648]\n",
      "   [0.23529412 0.16862746 0.19215687]\n",
      "   [0.20392157 0.14509805 0.16470589]\n",
      "   ...\n",
      "   [0.37254903 0.32156864 0.2901961 ]\n",
      "   [0.4627451  0.4117647  0.38039216]\n",
      "   [0.43137255 0.39215687 0.3529412 ]]\n",
      "\n",
      "  [[0.4117647  0.34509805 0.36862746]\n",
      "   [0.30980393 0.24313726 0.26666668]\n",
      "   [0.19215687 0.1254902  0.14901961]\n",
      "   ...\n",
      "   [0.33333334 0.27450982 0.24705882]\n",
      "   [0.42745098 0.3764706  0.34117648]\n",
      "   [0.49803922 0.45882353 0.41960785]]\n",
      "\n",
      "  [[0.43529412 0.36078432 0.3764706 ]\n",
      "   [0.3647059  0.3019608  0.3137255 ]\n",
      "   [0.22352941 0.16078432 0.17254902]\n",
      "   ...\n",
      "   [0.34901962 0.2901961  0.2627451 ]\n",
      "   [0.4117647  0.36078432 0.3254902 ]\n",
      "   [0.47843137 0.43137255 0.38431373]]\n",
      "\n",
      "  ...\n",
      "\n",
      "  [[0.43137255 0.3529412  0.39607844]\n",
      "   [0.49411765 0.41568628 0.45882353]\n",
      "   [0.45490196 0.3764706  0.41960785]\n",
      "   ...\n",
      "   [0.29411766 0.23137255 0.27058825]\n",
      "   [0.30588236 0.24313726 0.28235295]\n",
      "   [0.27450982 0.21176471 0.2509804 ]]\n",
      "\n",
      "  [[0.40784314 0.32941177 0.37254903]\n",
      "   [0.44705883 0.36862746 0.4117647 ]\n",
      "   [0.4        0.32156864 0.3647059 ]\n",
      "   ...\n",
      "   [0.32941177 0.26666668 0.30588236]\n",
      "   [0.3372549  0.27450982 0.3137255 ]\n",
      "   [0.32156864 0.25882354 0.29803923]]\n",
      "\n",
      "  [[0.36078432 0.28235295 0.3254902 ]\n",
      "   [0.42745098 0.34901962 0.39215687]\n",
      "   [0.39607844 0.31764707 0.36078432]\n",
      "   ...\n",
      "   [0.3254902  0.2627451  0.3019608 ]\n",
      "   [0.3372549  0.27450982 0.3137255 ]\n",
      "   [0.33333334 0.27058825 0.30980393]]]\n",
      "\n",
      "\n",
      " [[[0.45882353 0.4117647  0.49019608]\n",
      "   [0.76862746 0.72156864 0.8       ]\n",
      "   [0.1254902  0.07843138 0.16470589]\n",
      "   ...\n",
      "   [0.22745098 0.19607843 0.24705882]\n",
      "   [0.17254902 0.14901961 0.19607843]\n",
      "   [0.2627451  0.23921569 0.28627452]]\n",
      "\n",
      "  [[0.40784314 0.36078432 0.4392157 ]\n",
      "   [0.29803923 0.2509804  0.32941177]\n",
      "   [0.4117647  0.3647059  0.4509804 ]\n",
      "   ...\n",
      "   [0.21960784 0.1882353  0.23921569]\n",
      "   [0.21960784 0.19607843 0.24313726]\n",
      "   [0.21176471 0.1882353  0.23529412]]\n",
      "\n",
      "  [[0.3647059  0.32156864 0.3882353 ]\n",
      "   [0.3764706  0.33333334 0.4       ]\n",
      "   [0.34509805 0.29803923 0.3764706 ]\n",
      "   ...\n",
      "   [0.22745098 0.19607843 0.24705882]\n",
      "   [0.20392157 0.18039216 0.22745098]\n",
      "   [0.23137255 0.20784314 0.25490198]]\n",
      "\n",
      "  ...\n",
      "\n",
      "  [[0.39215687 0.37254903 0.4       ]\n",
      "   [0.3529412  0.33333334 0.36078432]\n",
      "   [0.32156864 0.3019608  0.32941177]\n",
      "   ...\n",
      "   [0.25882354 0.25490198 0.27450982]\n",
      "   [0.24313726 0.24313726 0.2509804 ]\n",
      "   [0.31764707 0.31764707 0.3254902 ]]\n",
      "\n",
      "  [[0.29803923 0.28627452 0.32156864]\n",
      "   [0.34117648 0.32941177 0.3647059 ]\n",
      "   [0.28235295 0.2627451  0.2901961 ]\n",
      "   ...\n",
      "   [0.34117648 0.34117648 0.34901962]\n",
      "   [0.38039216 0.38431373 0.39215687]\n",
      "   [0.30980393 0.3137255  0.32156864]]\n",
      "\n",
      "  [[0.25882354 0.24705882 0.28235295]\n",
      "   [0.3372549  0.3254902  0.36078432]\n",
      "   [0.19215687 0.17254902 0.2       ]\n",
      "   ...\n",
      "   [0.39215687 0.39215687 0.4       ]\n",
      "   [0.10588235 0.10980392 0.11764706]\n",
      "   [0.34901962 0.3529412  0.36078432]]]], shape=(2, 32, 32, 3), dtype=float32)\n",
      "tf.Tensor(\n",
      "[[0 1]\n",
      " [0 1]], shape=(2, 2), dtype=int32)\n"
     ]
    }
   ],
   "source": [
    "mini_train_ds = build_training_dataset(X_train[:5], y_train[:5], batch_size=2)\n",
    "# Fetch and print the first batch of 2 images\n",
    "for images, labels in mini_train_ds.take(1):\n",
    "    print(images)\n",
    "    print(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "trained-transcription",
    "outputId": "34994e32-0d93-47e0-df80-ea8d8aaa5c45"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "flatten (Flatten)            (None, 3072)              0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 256)               786688    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                16448     \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 2)                 130       \n",
      "=================================================================\n",
      "Total params: 803,266\n",
      "Trainable params: 803,266\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = tf.keras.Sequential()\n",
    "# Initially the orginal code from the site did not have the input_shape\n",
    "# So, model.fit was not working, then I added the details of input_shape\n",
    "# To the Flatten API as shown below, input_shape and the model.fit works!!!!\n",
    "# input values are 32*32*3 = 3072 entries of one single vector (input image)\n",
    "model.add(tf.keras.layers.Flatten(input_shape = (32,32,3)))\n",
    "\n",
    "# 3072*256 = 786432 weights + 256 biases = 786688\n",
    "model.add(tf.keras.layers.Dense(256, activation='relu'))\n",
    "\n",
    "# 256*64 = 16384 weights + 64 biases = 16448\n",
    "model.add(tf.keras.layers.Dense(64, activation='relu'))\n",
    "\n",
    "# Dropout is 20% of inputs from the above layer is randomly ignored while\n",
    "# performing training and those nodes are also not conisered for updating the\n",
    "# weights on back propagation to avoid over fitting, by makeing the adjacent\n",
    "# neurons to become independent of each other in learning about the features.\n",
    "model.add(tf.keras.layers.Dropout(0.2)) # This is added to improve learning\n",
    "\n",
    "# 64 * 2 = 128 weights + 2 biases = 130 params\n",
    "model.add(tf.keras.layers.Dense(2, activation='softmax')) # not working act=sigmoid, softmax\n",
    "\n",
    "#model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', \n",
    "              metrics=['accuracy'])\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "rational-ontario",
    "outputId": "f5f1e14e-d715-47b8-d2f5-9f2ababa3e0a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n",
      "400/400 [==============================] - 38s 31ms/step - loss: 0.5997 - accuracy: 0.7280 - val_loss: 0.3745 - val_accuracy: 0.8060\n",
      "Epoch 2/15\n",
      "400/400 [==============================] - 9s 24ms/step - loss: 0.3814 - accuracy: 0.8219 - val_loss: 0.4066 - val_accuracy: 0.7909\n",
      "Epoch 3/15\n",
      "400/400 [==============================] - 11s 27ms/step - loss: 0.3417 - accuracy: 0.8472 - val_loss: 0.3280 - val_accuracy: 0.8400\n",
      "Epoch 4/15\n",
      "400/400 [==============================] - 11s 26ms/step - loss: 0.3290 - accuracy: 0.8589 - val_loss: 0.2888 - val_accuracy: 0.8826\n",
      "Epoch 5/15\n",
      "400/400 [==============================] - 13s 32ms/step - loss: 0.3108 - accuracy: 0.8664 - val_loss: 0.3071 - val_accuracy: 0.8626\n",
      "Epoch 6/15\n",
      "400/400 [==============================] - 13s 34ms/step - loss: 0.3124 - accuracy: 0.8658 - val_loss: 0.2803 - val_accuracy: 0.8866\n",
      "Epoch 7/15\n",
      "400/400 [==============================] - 14s 34ms/step - loss: 0.2733 - accuracy: 0.8858 - val_loss: 0.3570 - val_accuracy: 0.8031\n",
      "Epoch 8/15\n",
      "400/400 [==============================] - 13s 32ms/step - loss: 0.2583 - accuracy: 0.8950 - val_loss: 0.2512 - val_accuracy: 0.8969\n",
      "Epoch 9/15\n",
      "400/400 [==============================] - 12s 30ms/step - loss: 0.2735 - accuracy: 0.8855 - val_loss: 0.3981 - val_accuracy: 0.8060\n",
      "Epoch 10/15\n",
      "400/400 [==============================] - 12s 30ms/step - loss: 0.2697 - accuracy: 0.8895 - val_loss: 0.2607 - val_accuracy: 0.8837\n",
      "Epoch 11/15\n",
      "400/400 [==============================] - 13s 32ms/step - loss: 0.2979 - accuracy: 0.8693 - val_loss: 0.3313 - val_accuracy: 0.8457\n",
      "Epoch 12/15\n",
      "400/400 [==============================] - 12s 30ms/step - loss: 0.2809 - accuracy: 0.8780 - val_loss: 0.2731 - val_accuracy: 0.8831\n",
      "Epoch 13/15\n",
      "400/400 [==============================] - 11s 28ms/step - loss: 0.2530 - accuracy: 0.8933 - val_loss: 0.2593 - val_accuracy: 0.8894\n",
      "Epoch 14/15\n",
      "400/400 [==============================] - 11s 28ms/step - loss: 0.2558 - accuracy: 0.8967 - val_loss: 0.2596 - val_accuracy: 0.8903\n",
      "Epoch 15/15\n",
      "400/400 [==============================] - 12s 30ms/step - loss: 0.2461 - accuracy: 0.9006 - val_loss: 0.2707 - val_accuracy: 0.8851\n"
     ]
    }
   ],
   "source": [
    "# Changed the ecpochs from 20 to 15 since the loss stabilises\n",
    "history = model.fit(train_ds, epochs=15, steps_per_epoch=400, \n",
    "                    validation_data=validation_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "twenty-stuart",
    "outputId": "f1e1532c-a14c-4e57-a88d-7876d5a9cc6a"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_accuracies_and_losses(history):\n",
    "    plt.title('Accuracy')\n",
    "    plt.plot(history.history['accuracy'])\n",
    "    plt.plot(history.history['val_accuracy'])\n",
    "    plt.legend(['training', 'validation'], loc='upper left')\n",
    "    plt.show()\n",
    "    \n",
    "    plt.title('Cross-entropy loss')\n",
    "    plt.plot(history.history['loss'])\n",
    "    plt.plot(history.history['val_loss'])\n",
    "    plt.legend(['training', 'validation'], loc='upper left')\n",
    "    plt.show()\n",
    "\n",
    "plot_accuracies_and_losses(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "dedicated-queen",
    "outputId": "0b0c7e74-aab4-4c29-e010-5d16278c949e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4000\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['000940378805c44108d287872b2f04ce.jpg',\n",
       " '0017242f54ececa4512b4d7937d1e21e.jpg',\n",
       " '001ee6d8564003107853118ab87df407.jpg',\n",
       " '002e175c3c1e060769475f52182583d0.jpg',\n",
       " '0036e44a7e8f7218e9bc7bf8137e4943.jpg']"
      ]
     },
     "execution_count": 14,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_dir = 'dataFromKaggle/aerial-cactus-identification/test/'\n",
    "test_imgs = listdir(test_dir)\n",
    "print(len(test_imgs))\n",
    "test_imgs[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "capable-lithuania",
    "outputId": "f1f2949d-8cfd-42ad-ebf9-b66fbcf62b87"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[[0.22352941, 0.23137255, 0.21176471],\n",
       "         [0.20784314, 0.21568628, 0.19607843],\n",
       "         [0.10588235, 0.11372549, 0.09411765],\n",
       "         ...,\n",
       "         [0.21960784, 0.17254902, 0.21176471],\n",
       "         [0.2       , 0.15294118, 0.19215687],\n",
       "         [0.23529412, 0.1882353 , 0.22745098]],\n",
       "\n",
       "        [[0.20392157, 0.21176471, 0.19215687],\n",
       "         [0.2       , 0.20784314, 0.1882353 ],\n",
       "         [0.11764706, 0.1254902 , 0.10588235],\n",
       "         ...,\n",
       "         [0.23137255, 0.18431373, 0.22352941],\n",
       "         [0.20392157, 0.15686275, 0.19607843],\n",
       "         [0.2627451 , 0.21568628, 0.25490198]],\n",
       "\n",
       "        [[0.38039216, 0.3882353 , 0.36862746],\n",
       "         [0.34509805, 0.3529412 , 0.33333334],\n",
       "         [0.23137255, 0.23921569, 0.21960784],\n",
       "         ...,\n",
       "         [0.24705882, 0.2       , 0.23921569],\n",
       "         [0.22352941, 0.1764706 , 0.21568628],\n",
       "         [0.25882354, 0.21176471, 0.2509804 ]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[0.32156864, 0.29411766, 0.3254902 ],\n",
       "         [0.3137255 , 0.28627452, 0.31764707],\n",
       "         [0.32156864, 0.3019608 , 0.32941177],\n",
       "         ...,\n",
       "         [0.43137255, 0.3882353 , 0.41960785],\n",
       "         [0.5176471 , 0.4745098 , 0.5058824 ],\n",
       "         [0.50980395, 0.46666667, 0.49803922]],\n",
       "\n",
       "        [[0.16470589, 0.13725491, 0.16862746],\n",
       "         [0.17254902, 0.14509805, 0.1764706 ],\n",
       "         [0.21568628, 0.1882353 , 0.21960784],\n",
       "         ...,\n",
       "         [0.43529412, 0.39215687, 0.42352942],\n",
       "         [0.5176471 , 0.4745098 , 0.5058824 ],\n",
       "         [0.4745098 , 0.43137255, 0.4627451 ]],\n",
       "\n",
       "        [[0.18431373, 0.15686275, 0.1882353 ],\n",
       "         [0.18431373, 0.15686275, 0.1882353 ],\n",
       "         [0.19607843, 0.16862746, 0.2       ],\n",
       "         ...,\n",
       "         [0.44313726, 0.4       , 0.43137255],\n",
       "         [0.49803922, 0.45490196, 0.4862745 ],\n",
       "         [0.41960785, 0.3764706 , 0.40784314]]],\n",
       "\n",
       "\n",
       "       [[[0.30980393, 0.2784314 , 0.27058825],\n",
       "         [0.32941177, 0.29803923, 0.2901961 ],\n",
       "         [0.32941177, 0.29803923, 0.2901961 ],\n",
       "         ...,\n",
       "         [0.2901961 , 0.29411766, 0.3019608 ],\n",
       "         [0.4392157 , 0.4392157 , 0.44705883],\n",
       "         [0.4392157 , 0.4392157 , 0.44705883]],\n",
       "\n",
       "        [[0.29803923, 0.26666668, 0.25882354],\n",
       "         [0.28627452, 0.25490198, 0.24705882],\n",
       "         [0.39607844, 0.3647059 , 0.35686275],\n",
       "         ...,\n",
       "         [0.16862746, 0.17254902, 0.18039216],\n",
       "         [0.3019608 , 0.3019608 , 0.30980393],\n",
       "         [0.4       , 0.4       , 0.40784314]],\n",
       "\n",
       "        [[0.21568628, 0.19607843, 0.18431373],\n",
       "         [0.23921569, 0.21960784, 0.20784314],\n",
       "         [0.34509805, 0.3254902 , 0.3137255 ],\n",
       "         ...,\n",
       "         [0.33333334, 0.33333334, 0.34117648],\n",
       "         [0.14117648, 0.13333334, 0.14509805],\n",
       "         [0.19215687, 0.18431373, 0.19607843]],\n",
       "\n",
       "        ...,\n",
       "\n",
       "        [[0.30980393, 0.3137255 , 0.29411766],\n",
       "         [0.3529412 , 0.35686275, 0.3372549 ],\n",
       "         [0.3529412 , 0.35686275, 0.3372549 ],\n",
       "         ...,\n",
       "         [0.22352941, 0.17254902, 0.2       ],\n",
       "         [0.23529412, 0.18431373, 0.21176471],\n",
       "         [0.21960784, 0.16862746, 0.19607843]],\n",
       "\n",
       "        [[0.36078432, 0.3647059 , 0.34509805],\n",
       "         [0.35686275, 0.36078432, 0.34117648],\n",
       "         [0.34901962, 0.3529412 , 0.33333334],\n",
       "         ...,\n",
       "         [0.4627451 , 0.4117647 , 0.4392157 ],\n",
       "         [0.21960784, 0.16862746, 0.19607843],\n",
       "         [0.29803923, 0.24705882, 0.27450982]],\n",
       "\n",
       "        [[0.29803923, 0.3019608 , 0.28235295],\n",
       "         [0.2627451 , 0.26666668, 0.24705882],\n",
       "         [0.28627452, 0.2901961 , 0.27058825],\n",
       "         ...,\n",
       "         [0.4509804 , 0.4       , 0.42745098],\n",
       "         [0.21960784, 0.16862746, 0.19607843],\n",
       "         [0.24313726, 0.19215687, 0.21960784]]]], dtype=float32)"
      ]
     },
     "execution_count": 15,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def path_to_numpy_array(path):\n",
    "    tensor = img_to_tensor(path)\n",
    "    array = tensor.numpy()\n",
    "    return array\n",
    "\n",
    "test_image_paths = [test_dir + ti for ti in test_imgs]\n",
    "test_instances = np.asarray([path_to_numpy_array(tip) for tip in test_image_paths])\n",
    "\n",
    "test_instances[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "english-joseph",
    "outputId": "b12ed25d-4d99-42bf-a6ea-32f3cae98594"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n",
      "<class 'list'>\n",
      "4000\n",
      "4000\n",
      "000940378805c44108d287872b2f04ce.jpg\n",
      "0017242f54ececa4512b4d7937d1e21e.jpg\n",
      "001ee6d8564003107853118ab87df407.jpg\n",
      "002e175c3c1e060769475f52182583d0.jpg\n",
      "0036e44a7e8f7218e9bc7bf8137e4943.jpg\n",
      "0038ed6f7417b8275b393307f66cb1e3.jpg\n",
      "0059dfa41de4441fc23b9a4b98ba99cb.jpg\n",
      "0061bd302d735db829418cb90b4e6040.jpg\n",
      "00754573d8c7223e73da553dd43780bd.jpg\n",
      "008d9fce71c8e03a4c9cdd15ea908573.jpg\n",
      "[0.00667286 0.99332714]\n",
      "[0.00906924 0.99093074]\n",
      "[0.9583354  0.04166457]\n",
      "[0.4576697 0.5423303]\n",
      "[0.01286176 0.9871383 ]\n",
      "[0.0012212  0.99877876]\n",
      "[0.00929768 0.9907023 ]\n",
      "[0.05308602 0.946914  ]\n",
      "[0.00239801 0.99760205]\n",
      "[0.88328683 0.1167131 ]\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(test_instances)\n",
    "print(type(predictions))\n",
    "print(type(test_imgs))\n",
    "print(len(predictions))\n",
    "print(len(test_imgs))\n",
    "for i in range(10):\n",
    "    print(test_imgs[i])\n",
    "for i in range(10):\n",
    "    print(predictions[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "geographic-holiday",
    "outputId": "261bc93b-1b5c-4e54-fec8-cdcd8c76f850"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>not_a_cactus</th>\n",
       "      <th>has_cactus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>000940378805c44108d287872b2f04ce.jpg</td>\n",
       "      <td>0.006673</td>\n",
       "      <td>0.993327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0017242f54ececa4512b4d7937d1e21e.jpg</td>\n",
       "      <td>0.009069</td>\n",
       "      <td>0.990931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>001ee6d8564003107853118ab87df407.jpg</td>\n",
       "      <td>0.958335</td>\n",
       "      <td>0.041665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>002e175c3c1e060769475f52182583d0.jpg</td>\n",
       "      <td>0.457670</td>\n",
       "      <td>0.542330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0036e44a7e8f7218e9bc7bf8137e4943.jpg</td>\n",
       "      <td>0.012862</td>\n",
       "      <td>0.987138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0038ed6f7417b8275b393307f66cb1e3.jpg</td>\n",
       "      <td>0.001221</td>\n",
       "      <td>0.998779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0059dfa41de4441fc23b9a4b98ba99cb.jpg</td>\n",
       "      <td>0.009298</td>\n",
       "      <td>0.990702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0061bd302d735db829418cb90b4e6040.jpg</td>\n",
       "      <td>0.053086</td>\n",
       "      <td>0.946914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>00754573d8c7223e73da553dd43780bd.jpg</td>\n",
       "      <td>0.002398</td>\n",
       "      <td>0.997602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>008d9fce71c8e03a4c9cdd15ea908573.jpg</td>\n",
       "      <td>0.883287</td>\n",
       "      <td>0.116713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>00a36d4d6d152404670276fc983273bc.jpg</td>\n",
       "      <td>0.005148</td>\n",
       "      <td>0.994852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>00b706122b87e0fa275ff59e39d4d94b.jpg</td>\n",
       "      <td>0.003423</td>\n",
       "      <td>0.996576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>00c054da839d5518e59790f7d867f317.jpg</td>\n",
       "      <td>0.219196</td>\n",
       "      <td>0.780804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>00c88441b0510cdb3a6e9b3fa7b632af.jpg</td>\n",
       "      <td>0.920745</td>\n",
       "      <td>0.079255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>00e1e29865202c8ca715b0f14848d577.jpg</td>\n",
       "      <td>0.905205</td>\n",
       "      <td>0.094795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>00e20b0e6d779e7c36cea29710eed88e.jpg</td>\n",
       "      <td>0.196202</td>\n",
       "      <td>0.803798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>00feb0be23ac80f397a55b0ed8502def.jpg</td>\n",
       "      <td>0.001203</td>\n",
       "      <td>0.998797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>01185a1a81bb8041fa5e9c9ada9374eb.jpg</td>\n",
       "      <td>0.091837</td>\n",
       "      <td>0.908163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0124398c0fcf6d1c92d9800337636b4e.jpg</td>\n",
       "      <td>0.003635</td>\n",
       "      <td>0.996365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0127044dfc88dfaed0118c8764909800.jpg</td>\n",
       "      <td>0.567429</td>\n",
       "      <td>0.432571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      id  not_a_cactus  has_cactus\n",
       "0   000940378805c44108d287872b2f04ce.jpg      0.006673    0.993327\n",
       "1   0017242f54ececa4512b4d7937d1e21e.jpg      0.009069    0.990931\n",
       "2   001ee6d8564003107853118ab87df407.jpg      0.958335    0.041665\n",
       "3   002e175c3c1e060769475f52182583d0.jpg      0.457670    0.542330\n",
       "4   0036e44a7e8f7218e9bc7bf8137e4943.jpg      0.012862    0.987138\n",
       "5   0038ed6f7417b8275b393307f66cb1e3.jpg      0.001221    0.998779\n",
       "6   0059dfa41de4441fc23b9a4b98ba99cb.jpg      0.009298    0.990702\n",
       "7   0061bd302d735db829418cb90b4e6040.jpg      0.053086    0.946914\n",
       "8   00754573d8c7223e73da553dd43780bd.jpg      0.002398    0.997602\n",
       "9   008d9fce71c8e03a4c9cdd15ea908573.jpg      0.883287    0.116713\n",
       "10  00a36d4d6d152404670276fc983273bc.jpg      0.005148    0.994852\n",
       "11  00b706122b87e0fa275ff59e39d4d94b.jpg      0.003423    0.996576\n",
       "12  00c054da839d5518e59790f7d867f317.jpg      0.219196    0.780804\n",
       "13  00c88441b0510cdb3a6e9b3fa7b632af.jpg      0.920745    0.079255\n",
       "14  00e1e29865202c8ca715b0f14848d577.jpg      0.905205    0.094795\n",
       "15  00e20b0e6d779e7c36cea29710eed88e.jpg      0.196202    0.803798\n",
       "16  00feb0be23ac80f397a55b0ed8502def.jpg      0.001203    0.998797\n",
       "17  01185a1a81bb8041fa5e9c9ada9374eb.jpg      0.091837    0.908163\n",
       "18  0124398c0fcf6d1c92d9800337636b4e.jpg      0.003635    0.996365\n",
       "19  0127044dfc88dfaed0118c8764909800.jpg      0.567429    0.432571"
      ]
     },
     "execution_count": 17,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#submission_data = pd.DataFrame({'id': test_imgs, 'has_cactus': predictions.flatten()})\n",
    "predictions_first_elm = []\n",
    "predictions_second_elm = []\n",
    "\n",
    "for lstObj in predictions:\n",
    "    predictions_first_elm.append(lstObj[0])\n",
    "    predictions_second_elm.append(lstObj[1])\n",
    "    \n",
    "\n",
    "submission_data = pd.DataFrame({'id': test_imgs, \n",
    "                                'not_a_cactus' : predictions_first_elm, \\\n",
    "                                'has_cactus': predictions_second_elm})\n",
    "\n",
    "submission_data.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "hybrid-bahamas"
   },
   "outputs": [],
   "source": [
    "submission_data.to_csv('submission_seq.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "absent-butter",
    "outputId": "db9de6a8-dad5-4ff5-bcb6-8ca65c4c657d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total entries: 8000\n",
      "                                      not_a_cactus  has_cactus\n",
      "id                                                            \n",
      "000940378805c44108d287872b2f04ce.jpg      0.006673    0.993327\n",
      "0017242f54ececa4512b4d7937d1e21e.jpg      0.009069    0.990931\n",
      "001ee6d8564003107853118ab87df407.jpg      0.958335    0.041665\n",
      "002e175c3c1e060769475f52182583d0.jpg      0.457670    0.542330\n",
      "0036e44a7e8f7218e9bc7bf8137e4943.jpg      0.012862    0.987138\n",
      "0038ed6f7417b8275b393307f66cb1e3.jpg      0.001221    0.998779\n",
      "0059dfa41de4441fc23b9a4b98ba99cb.jpg      0.009298    0.990702\n",
      "0061bd302d735db829418cb90b4e6040.jpg      0.053086    0.946914\n",
      "00754573d8c7223e73da553dd43780bd.jpg      0.002398    0.997602\n",
      "008d9fce71c8e03a4c9cdd15ea908573.jpg      0.883287    0.116713\n"
     ]
    }
   ],
   "source": [
    "# Replaced the !head in the reference code with the following\n",
    "submission_df = pd.read_csv('submission_seq.csv',\n",
    "                              index_col='id')\n",
    "print('Total entries: ' + str(submission_df.size))\n",
    "print(submission_df.head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "appreciated-recorder",
    "outputId": "e0bb3b0c-d2f2-4399-d229-7ff48a7d8b2c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4000, 32, 32, 3)\n"
     ]
    }
   ],
   "source": [
    "X_test = test_instances\n",
    "print(X_test.shape)\n",
    "predictions = model.predict(X_test)\n",
    "#for i in range(len(predictions)):\n",
    "#    print(predictions[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "white-halifax"
   },
   "outputs": [],
   "source": [
    "# It save text file with Hex numbers in HDF5 format in the current dir\n",
    "model.save('CactusClassifySeqModel.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "preceding-runner",
    "outputId": "25b08d7e-1d84-4abe-c696-d14154336a73"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.00667286 0.99332714]\n",
      " [0.00906924 0.99093074]\n",
      " [0.9583354  0.04166457]\n",
      " ...\n",
      " [0.00123443 0.9987656 ]\n",
      " [0.01582965 0.9841703 ]\n",
      " [0.14016184 0.85983807]]\n",
      "[0.00667286 0.99332714]\n",
      "[0.9583354  0.04166457]\n"
     ]
    }
   ],
   "source": [
    "predictions = model.predict(X_test)\n",
    "print(predictions)\n",
    "print(predictions[0])\n",
    "print(predictions[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "smoking-semiconductor",
    "outputId": "c77adcf8-3d91-4804-877d-5dac70b00b61"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<generator object representative_dataset at 0x000001305F4617C8>\n"
     ]
    }
   ],
   "source": [
    "def representative_dataset():\n",
    "    for _ in range(100):\n",
    "      data =  X_test\n",
    "      yield [data.astype(np.float32)]\n",
    "        \n",
    "print(representative_dataset())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "cooperative-aberdeen",
    "outputId": "41aaf32d-aedd-42e0-e5e6-826f90867d35"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: saved_cactus_seq_model_keras_dir\\assets\n"
     ]
    }
   ],
   "source": [
    "tf.saved_model.save(model, \"saved_cactus_seq_model_keras_dir\")\n",
    "converter = tf.lite.TFLiteConverter.from_saved_model(\"saved_cactus_seq_model_keras_dir\")\n",
    "converter.optimizations = [tf.lite.Optimize.DEFAULT]\n",
    "converter.representative_dataset = representative_dataset\n",
    "tflite_model = converter.convert()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "respective-steam"
   },
   "outputs": [],
   "source": [
    "with open('CactusClassifySeqModel.tflite', 'wb') as f:\n",
    "  f.write(tflite_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "fiscal-situation"
   },
   "outputs": [],
   "source": [
    "interpreter = tf.lite.Interpreter(model_path=\"CactusClassifySeqModel.tflite\")\n",
    "interpreter.allocate_tensors()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "threatened-favorite",
    "outputId": "7f174cd7-95cb-41d3-c0bc-de36674dc758"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input_details:\n",
      " [{'name': 'serving_default_flatten_input:0', 'index': 13, 'shape': array([ 1, 32, 32,  3]), 'shape_signature': array([-1, 32, 32,  3]), 'dtype': <class 'numpy.float32'>, 'quantization': (0.0, 0), 'quantization_parameters': {'scales': array([], dtype=float32), 'zero_points': array([], dtype=int32), 'quantized_dimension': 0}, 'sparsity_parameters': {}}]\n",
      "output_details:\n",
      " [{'name': 'StatefulPartitionedCall:0', 'index': 14, 'shape': array([1, 2]), 'shape_signature': array([-1,  2]), 'dtype': <class 'numpy.float32'>, 'quantization': (0.0, 0), 'quantization_parameters': {'scales': array([], dtype=float32), 'zero_points': array([], dtype=int32), 'quantized_dimension': 0}, 'sparsity_parameters': {}}]\n"
     ]
    }
   ],
   "source": [
    "# Get input and output tensors.\n",
    "input_details = interpreter.get_input_details()\n",
    "output_details = interpreter.get_output_details()\n",
    "print('input_details:\\n', input_details)\n",
    "print('output_details:\\n', output_details)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "western-airplane",
    "outputId": "99dd3c51-ac50-402a-b4fe-c43b5a18e0ad"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 1 32 32  3]\n",
      "[array([[[0.22352941, 0.23137255, 0.21176471],\n",
      "        [0.20784314, 0.21568628, 0.19607843],\n",
      "        [0.10588235, 0.11372549, 0.09411765],\n",
      "        ...,\n",
      "        [0.21960784, 0.17254902, 0.21176471],\n",
      "        [0.2       , 0.15294118, 0.19215687],\n",
      "        [0.23529412, 0.1882353 , 0.22745098]],\n",
      "\n",
      "       [[0.20392157, 0.21176471, 0.19215687],\n",
      "        [0.2       , 0.20784314, 0.1882353 ],\n",
      "        [0.11764706, 0.1254902 , 0.10588235],\n",
      "        ...,\n",
      "        [0.23137255, 0.18431373, 0.22352941],\n",
      "        [0.20392157, 0.15686275, 0.19607843],\n",
      "        [0.2627451 , 0.21568628, 0.25490198]],\n",
      "\n",
      "       [[0.38039216, 0.3882353 , 0.36862746],\n",
      "        [0.34509805, 0.3529412 , 0.33333334],\n",
      "        [0.23137255, 0.23921569, 0.21960784],\n",
      "        ...,\n",
      "        [0.24705882, 0.2       , 0.23921569],\n",
      "        [0.22352941, 0.1764706 , 0.21568628],\n",
      "        [0.25882354, 0.21176471, 0.2509804 ]],\n",
      "\n",
      "       ...,\n",
      "\n",
      "       [[0.32156864, 0.29411766, 0.3254902 ],\n",
      "        [0.3137255 , 0.28627452, 0.31764707],\n",
      "        [0.32156864, 0.3019608 , 0.32941177],\n",
      "        ...,\n",
      "        [0.43137255, 0.3882353 , 0.41960785],\n",
      "        [0.5176471 , 0.4745098 , 0.5058824 ],\n",
      "        [0.50980395, 0.46666667, 0.49803922]],\n",
      "\n",
      "       [[0.16470589, 0.13725491, 0.16862746],\n",
      "        [0.17254902, 0.14509805, 0.1764706 ],\n",
      "        [0.21568628, 0.1882353 , 0.21960784],\n",
      "        ...,\n",
      "        [0.43529412, 0.39215687, 0.42352942],\n",
      "        [0.5176471 , 0.4745098 , 0.5058824 ],\n",
      "        [0.4745098 , 0.43137255, 0.4627451 ]],\n",
      "\n",
      "       [[0.18431373, 0.15686275, 0.1882353 ],\n",
      "        [0.18431373, 0.15686275, 0.1882353 ],\n",
      "        [0.19607843, 0.16862746, 0.2       ],\n",
      "        ...,\n",
      "        [0.44313726, 0.4       , 0.43137255],\n",
      "        [0.49803922, 0.45490196, 0.4862745 ],\n",
      "        [0.41960785, 0.3764706 , 0.40784314]]], dtype=float32)]\n",
      "[[0.0078125 0.9921875]]\n"
     ]
    }
   ],
   "source": [
    "# Test the model on random input data.\n",
    "input_shape = input_details[0]['shape']\n",
    "print(input_shape)\n",
    "input_data = [X_test[0]]\n",
    "print(input_data)\n",
    "input_data = np.array(input_data, dtype=np.float32)\n",
    "interpreter.set_tensor(input_details[0]['index'], input_data)\n",
    "\n",
    "interpreter.invoke()\n",
    "\n",
    "# The function `get_tensor()` returns a copy of the tensor data.\n",
    "# Use `tensor()` in order to get a pointer to the tensor.\n",
    "output_data = interpreter.get_tensor(output_details[0]['index'])\n",
    "print(output_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "atmospheric-supply"
   },
   "outputs": [],
   "source": [
    "# Function to convert some hex values into an array for C programming\n",
    "def hex_to_c_array(hex_data, var_name):\n",
    "    c_str = \"\"\n",
    "\n",
    "    # Create header guard\n",
    "    c_str += '#ifndef ' + var_name.upper() + '_H\\n'\n",
    "    c_str += \"#define \" + var_name.upper() + '_H\\n\\n'\n",
    "\n",
    "    # Add array length at the top of the file\n",
    "    c_str += '\\nconst unsigned int ' + var_name + '_len = ' + str(len(hex_data)) + ';\\n'\n",
    "\n",
    "    # Declare C variable\n",
    "    c_str += 'alignas(8) const unsigned char ' + var_name + '[] = {'\n",
    "    hex_array = []\n",
    "    for i, val in enumerate(hex_data):\n",
    "        # Construct string from hex\n",
    "        hex_str = format(val, '#04x')\n",
    "\n",
    "        # Add formating so each line stays within 80 characters\n",
    "        if (i + 1) < len(hex_data):\n",
    "          hex_str += ','\n",
    "        if (i + 1) % 12 == 0:\n",
    "          hex_str += '\\n'\n",
    "        hex_array.append(hex_str)\n",
    "\n",
    "    # Add closing brace\n",
    "    c_str += '\\n' + format(''.join(hex_array)) + '\\n};\\n\\n'\n",
    "\n",
    "    # Close out header guard\n",
    "    c_str += '#endif //' + var_name.upper() + '_H'\n",
    "\n",
    "    return c_str"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "occasional-retail"
   },
   "outputs": [],
   "source": [
    "# Write TFLite model to a C source (or header) file\n",
    "with open(\"cactus_seq_model_esp32\" + '.h', 'w') as file:\n",
    "  file.write(hex_to_c_array(tflite_model, \"cactus_seq_model_esp32\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "guided-universe",
    "outputId": "db2e0d03-b2e6-4438-f1df-5c25d774a823"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.22352941\n",
      "0.23137255\n",
      "0.21176471\n",
      "0.41960785\n",
      "0.3764706\n",
      "0.40784314\n",
      "The last element\n",
      "0.40784314\n",
      "The last but one\n",
      "0.3764706\n",
      "32\n",
      "32\n",
      "3\n"
     ]
    }
   ],
   "source": [
    "print(X_test[0][0][0][0])\n",
    "print(X_test[0][0][0][1])\n",
    "print(X_test[0][0][0][2])\n",
    "\n",
    "print(X_test[0][31][31][0])\n",
    "print(X_test[0][31][31][1])\n",
    "print(X_test[0][31][31][2])\n",
    "\n",
    "print('The last element')\n",
    "print(X_test[0][31][31][2])\n",
    "print('The last but one')\n",
    "print(X_test[0][31][31][1])\n",
    "print(len(X_test[0]))\n",
    "print(len(X_test[0][0]))\n",
    "print(len(X_test[0][0][0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "funny-aquarium",
    "outputId": "8a1346bc-5ebb-4e5e-cf19-a96c0b0317f7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3 32 32\n"
     ]
    }
   ],
   "source": [
    "# The same indices will be valid for the first element of X_test[0] \n",
    "# The length values initialized here are the same for any element of X_test[]\n",
    "lenK = len(X_test[2][0][0])\n",
    "lenJ = len(X_test[2][0])\n",
    "lenI = len(X_test[2][0])\n",
    "print(lenK, lenJ, lenI)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "thrown-pipeline",
    "outputId": "30c96662-2314-4cf6-b44a-ec1bd155c788"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cactus_X_test0_data.h\n",
      "\n",
      " X_test0[]:\n",
      " {\n",
      " 0.22352941f, 0.23137255f, 0.21176471f, 0.20784314f, 0.21568628f, 0.19607843f, 0.105882354f, 0.11372549f, 0.09411765f, 0.17254902f, 0.18039216f, 0.16078432f, 0.27058825f, 0.2901961f, 0.26666668f, 0.43529412f, 0.45490196f, 0.43137255f, 0.41960785f, 0.4392157f, 0.42352942f, 0.3647059f, 0.38431373f, 0.36862746f, 0.25490198f, 0.27450982f, 0.25882354f, 0.17254902f, 0.19215687f, 0.1764706f, 0.25882354f, 0.26666668f, 0.25490198f, 0.28627452f, 0.29411766f, 0.28235295f, 0.32941177f, 0.32941177f, 0.32941177f, 0.34509805f, 0.34509805f, 0.34509805f, 0.32156864f, 0.32156864f, 0.32156864f, 0.1764706f, 0.1764706f, 0.1764706f, 0.18039216f, 0.16078432f, 0.1764706f, 0.16862746f, 0.14901961f, 0.16470589f, 0.3019608f, 0.28235295f, 0.29803923f, 0.30588236f, 0.28627452f, 0.3019608f, 0.23529412f, 0.21176471f, 0.22745098f, 0.26666668f, 0.24313726f, 0.25882354f, 0.21568628f, 0.1882353f, 0.21568628f, 0.29803923f, 0.27058825f, 0.29803923f, 0.20784314f, 0.16862746f, 0.2f, 0.21960784f, 0.18039216f, 0.21176471f, 0.17254902f, 0.13333334f, 0.16862746f, 0.16078432f, 0.12156863f, 0.15686275f, 0.24313726f, 0.19607843f, 0.23529412f, 0.21960784f, 0.17254902f, 0.21176471f, 0.2f, 0.15294118f, 0.19215687f, 0.23529412f, 0.1882353f, 0.22745098f, 0.20392157f, 0.21176471f, 0.19215687f, 0.2f, 0.20784314f, 0.1882353f, 0.11764706f, 0.1254902f, 0.105882354f, 0.16470589f, 0.17254902f, 0.15294118f, 0.27058825f, 0.2901961f, 0.26666668f, 0.41960785f, 0.4392157f, 0.41568628f, 0.42352942f, 0.44313726f, 0.42745098f, 0.36862746f, 0.3882353f, 0.37254903f, 0.23137255f, 0.2509804f, 0.23529412f, 0.18431373f, 0.20392157f, 0.1882353f, 0.25490198f, 0.2627451f, 0.2509804f, 0.2784314f, 0.28627452f, 0.27450982f, 0.35686275f, 0.35686275f, 0.35686275f, 0.34117648f, 0.34117648f, 0.34117648f, 0.2901961f, 0.2901961f, 0.2901961f, 0.1882353f, 0.1882353f, 0.1882353f, 0.1764706f, 0.15686275f, 0.17254902f, 0.1764706f, 0.15686275f, 0.17254902f, 0.28627452f, 0.26666668f, 0.28235295f, 0.30588236f, 0.28627452f, 0.3019608f, 0.2509804f, 0.22745098f, 0.24313726f, 0.2627451f, 0.23921569f, 0.25490198f, 0.24313726f, 0.21568628f, 0.24313726f, 0.27450982f, 0.24705882f, 0.27450982f, 0.21960784f, 0.18039216f, 0.21176471f, 0.21176471f, 0.17254902f, 0.20392157f, 0.16862746f, 0.12941177f, 0.16470589f, 0.16470589f, 0.1254902f, 0.16078432f, 0.20392157f, 0.15686275f, 0.19607843f, 0.23137255f, 0.18431373f, 0.22352941f, 0.20392157f, 0.15686275f, 0.19607843f, 0.2627451f, 0.21568628f, 0.25490198f, 0.38039216f, 0.3882353f, 0.36862746f, 0.34509805f, 0.3529412f, 0.33333334f, 0.23137255f, 0.23921569f, 0.21960784f, 0.1764706f, 0.18431373f, 0.16470589f, 0.21960784f, 0.23921569f, 0.22352941f, 0.29803923f, 0.31764707f, 0.3019608f, 0.32156864f, 0.34117648f, 0.3254902f, 0.27450982f, 0.29411766f, 0.2784314f, 0.20784314f, 0.22745098f, 0.21176471f, 0.20784314f, 0.22745098f, 0.21176471f, 0.2627451f, 0.27058825f, 0.26666668f, 0.25882354f, 0.26666668f, 0.2627451f, 0.3372549f, 0.3372549f, 0.3372549f, 0.2901961f, 0.2901961f, 0.2901961f, 0.22745098f, 0.22745098f, 0.22745098f, 0.18431373f, 0.18431373f, 0.18431373f, 0.14901961f, 0.12941177f, 0.14509805f, 0.15294118f, 0.13333334f, 0.14901961f, 0.24705882f, 0.22745098f, 0.24313726f, 0.3254902f, 0.30588236f, 0.32156864f, 0.3254902f, 0.3019608f, 0.31764707f, 0.2901961f, 0.26666668f, 0.28235295f, 0.2901961f, 0.2627451f, 0.2901961f, 0.23137255f, 0.20392157f, 0.23137255f, 0.1764706f, 0.13725491f, 0.16862746f, 0.16862746f, 0.12941177f, 0.16078432f, 0.19607843f, 0.15686275f, 0.1882353f, 0.1882353f, 0.14901961f, 0.18039216f, 0.20784314f, 0.16078432f, 0.2f, 0.24705882f, 0.2f, 0.23921569f, 0.22352941f, 0.1764706f, 0.21568628f, 0.25882354f, 0.21176471f, 0.2509804f, 0.30588236f, 0.3137255f, 0.29411766f, 0.30588236f, 0.3137255f, 0.29411766f, 0.2627451f, 0.27058825f, 0.2509804f, 0.20784314f, 0.21568628f, 0.19607843f, 0.20392157f, 0.22352941f, 0.20784314f, 0.16862746f, 0.1882353f, 0.17254902f, 0.16078432f, 0.18039216f, 0.16470589f, 0.11372549f, 0.13333334f, 0.11764706f, 0.09411765f, 0.11372549f, 0.09803922f, 0.14901961f, 0.16862746f, 0.15294118f, 0.2627451f, 0.27058825f, 0.26666668f, 0.28235295f, 0.2901961f, 0.28627452f, 0.35686275f, 0.35686275f, 0.35686275f, 0.31764707f, 0.31764707f, 0.31764707f, 0.27058825f, 0.27058825f, 0.27058825f, 0.23921569f, 0.23137255f, 0.23529412f, 0.21176471f, 0.19215687f, 0.20784314f, 0.2f, 0.18039216f, 0.19607843f, 0.25490198f, 0.23529412f, 0.2509804f, 0.3764706f, 0.35686275f, 0.37254903f, 0.39607844f, 0.37254903f, 0.3882353f, 0.30588236f, 0.28235295f, 0.29803923f, 0.3019608f, 0.27450982f, 0.3019608f, 0.18039216f, 0.15294118f, 0.18039216f, 0.105882354f, 0.06666667f, 0.09803922f, 0.1882353f, 0.14901961f, 0.18039216f, 0.2901961f, 0.2509804f, 0.28235295f, 0.28627452f, 0.24705882f, 0.2784314f, 0.2509804f, 0.20392157f, 0.24313726f, 0.2509804f, 0.20392157f, 0.24313726f, 0.29803923f, 0.2509804f, 0.2901961f, 0.28627452f, 0.23921569f, 0.2784314f, 0.2509804f, 0.25882354f, 0.24705882f, 0.27058825f, 0.2784314f, 0.26666668f, 0.27450982f, 0.28235295f, 0.27058825f, 0.23921569f, 0.24705882f, 0.23529412f, 0.21176471f, 0.23137255f, 0.21568628f, 0.11764706f, 0.13725491f, 0.12156863f, 0.10980392f, 0.12941177f, 0.11372549f, 0.09411765f, 0.11372549f, 0.09803922f, 0.18039216f, 0.19607843f, 0.19215687f, 0.16862746f, 0.18431373f, 0.18039216f, 0.24705882f, 0.25490198f, 0.2509804f, 0.23137255f, 0.23921569f, 0.23529412f, 0.29803923f, 0.29803923f, 0.29803923f, 0.3529412f, 0.3529412f, 0.3529412f, 0.40784314f, 0.4f, 0.40392157f, 0.38039216f, 0.37254903f, 0.3764706f, 0.2627451f, 0.24313726f, 0.25882354f, 0.2627451f, 0.24313726f, 0.25882354f, 0.3137255f, 0.29411766f, 0.30980393f, 0.44313726f, 0.42352942f, 0.4392157f, 0.45490196f, 0.43137255f, 0.44705883f, 0.31764707f, 0.29411766f, 0.30980393f, 0.29803923f, 0.27450982f, 0.2901961f, 0.1764706f, 0.15294118f, 0.16862746f, 0.21176471f, 0.17254902f, 0.20392157f, 0.49411765f, 0.45490196f, 0.4862745f, 0.627451f, 0.5882353f, 0.61960787f, 0.6f, 0.56078434f, 0.5921569f, 0.29411766f, 0.2509804f, 0.28235295f, 0.17254902f, 0.12941177f, 0.16078432f, 0.2509804f, 0.20784314f, 0.23921569f, 0.25882354f, 0.21568628f, 0.24705882f, 0.28627452f, 0.29411766f, 0.28235295f, 0.27450982f, 0.28235295f, 0.27058825f, 0.2509804f, 0.25882354f, 0.24705882f, 0.21568628f, 0.22352941f, 0.21176471f, 0.20784314f, 0.22745098f, 0.21176471f, 0.16862746f, 0.1882353f, 0.17254902f, 0.22352941f, 0.24313726f, 0.22745098f, 0.26666668f, 0.28627452f, 0.27058825f, 0.26666668f, 0.28235295f, 0.2784314f, 0.21568628f, 0.23137255f, 0.22745098f, 0.28627452f, 0.29411766f, 0.2901961f, 0.28235295f, 0.2901961f, 0.28627452f, 0.33333334f, 0.33333334f, 0.33333334f, 0.38039216f, 0.38039216f, 0.38039216f, 0.41960785f, 0.4117647f, 0.41568628f, 0.34509805f, 0.3372549f, 0.34117648f, 0.3372549f, 0.31764707f, 0.33333334f, 0.32941177f, 0.30980393f, 0.3254902f, 0.32941177f, 0.30980393f, 0.3254902f, 0.37254903f, 0.3529412f, 0.36862746f, 0.35686275f, 0.33333334f, 0.34901962f, 0.28627452f, 0.2627451f, 0.2784314f, 0.30980393f, 0.28627452f, 0.3019608f, 0.28235295f, 0.25882354f, 0.27450982f, 0.3764706f, 0.3372549f, 0.36862746f, 0.6431373f, 0.6039216f, 0.63529414f, 0.6313726f, 0.5921569f, 0.62352943f, 0.6392157f, 0.6f, 0.6313726f, 0.2509804f, 0.20784314f, 0.23921569f, 0.18039216f, 0.13725491f, 0.16862746f, 0.2509804f, 0.20784314f, 0.23921569f, 0.26666668f, 0.22352941f, 0.25490198f, 0.29411766f, 0.29411766f, 0.29411766f, 0.28235295f, 0.28235295f, 0.28235295f, 0.23529412f, 0.24313726f, 0.23921569f, 0.21176471f, 0.21960784f, 0.21568628f, 0.22745098f, 0.23529412f, 0.23137255f, 0.21176471f, 0.21960784f, 0.21568628f, 0.22352941f, 0.23921569f, 0.23529412f, 0.22745098f, 0.24313726f, 0.23921569f, 0.2901961f, 0.29803923f, 0.29411766f, 0.22745098f, 0.23529412f, 0.23137255f, 0.28627452f, 0.29411766f, 0.2901961f, 0.28627452f, 0.29411766f, 0.2901961f, 0.33333334f, 0.33333334f, 0.33333334f, 0.34901962f, 0.34901962f, 0.34901962f, 0.38431373f, 0.3764706f, 0.38039216f, 0.34509805f, 0.3372549f, 0.34117648f, 0.30980393f, 0.2901961f, 0.30588236f, 0.33333334f, 0.3137255f, 0.32941177f, 0.33333334f, 0.3137255f, 0.32941177f, 0.33333334f, 0.3137255f, 0.32941177f, 0.32156864f, 0.29803923f, 0.3137255f, 0.32156864f, 0.29803923f, 0.3137255f, 0.29803923f, 0.27450982f, 0.2901961f, 0.2901961f, 0.26666668f, 0.28235295f, 0.37254903f, 0.3372549f, 0.35686275f, 0.6156863f, 0.5803922f, 0.6f, 0.6313726f, 0.59607846f, 0.6156863f, 0.64705884f, 0.6117647f, 0.6313726f, 0.27058825f, 0.22745098f, 0.2509804f, 0.1764706f, 0.13333334f, 0.15686275f, 0.24313726f, 0.2f, 0.22352941f, 0.29803923f, 0.25490198f, 0.2784314f, 0.21960784f, 0.21960784f, 0.21960784f, 0.23529412f, 0.23529412f, 0.23529412f, 0.21176471f, 0.21960784f, 0.21568628f, 0.23137255f, 0.23921569f, 0.23529412f, 0.29411766f, 0.3019608f, 0.29803923f, 0.3254902f, 0.33333334f, 0.32941177f, 0.3137255f, 0.32941177f, 0.3254902f, 0.27450982f, 0.2901961f, 0.28627452f, 0.24705882f, 0.25490198f, 0.2509804f, 0.21960784f, 0.22745098f, 0.22352941f, 0.29803923f, 0.30588236f, 0.3019608f, 0.3137255f, 0.32156864f, 0.31764707f, 0.34509805f, 0.34509805f, 0.34509805f, 0.3019608f, 0.3019608f, 0.3019608f, 0.31764707f, 0.30980393f, 0.3137255f, 0.3254902f, 0.31764707f, 0.32156864f, 0.3254902f, 0.30588236f, 0.32156864f, 0.30588236f, 0.28627452f, 0.3019608f, 0.24313726f, 0.22352941f, 0.23921569f, 0.18039216f, 0.16078432f, 0.1764706f, 0.22745098f, 0.20392157f, 0.21960784f, 0.38039216f, 0.35686275f, 0.37254903f, 0.40784314f, 0.38431373f, 0.4f, 0.45882353f, 0.43529412f, 0.4509804f, 0.4392157f, 0.40392157f, 0.42352942f, 0.45490196f, 0.41960785f, 0.4392157f, 0.3764706f, 0.34117648f, 0.36078432f, 0.3647059f, 0.32941177f, 0.34901962f, 0.21568628f, 0.17254902f, 0.19607843f, 0.2f, 0.15686275f, 0.18039216f, 0.3019608f, 0.25882354f, 0.28235295f, 0.32941177f, 0.28627452f, 0.30980393f, 0.19215687f, 0.19215687f, 0.2f, 0.16470589f, 0.16470589f, 0.17254902f, 0.21568628f, 0.21960784f, 0.22745098f, 0.3019608f, 0.30588236f, 0.3137255f, 0.34509805f, 0.34901962f, 0.35686275f, 0.28627452f, 0.2901961f, 0.29803923f, 0.27450982f, 0.2784314f, 0.28627452f, 0.21960784f, 0.22352941f, 0.23137255f, 0.19607843f, 0.2f, 0.20784314f, 0.25490198f, 0.25882354f, 0.26666668f, 0.29411766f, 0.29411766f, 0.29411766f, 0.32941177f, 0.32941177f, 0.32941177f, 0.31764707f, 0.30980393f, 0.3137255f, 0.30588236f, 0.29803923f, 0.3019608f, 0.3254902f, 0.31764707f, 0.32156864f, 0.2901961f, 0.28235295f, 0.28627452f, 0.23137255f, 0.21176471f, 0.22745098f, 0.22352941f, 0.20392157f, 0.21960784f, 0.16078432f, 0.14117648f, 0.15686275f, 0.19607843f, 0.1764706f, 0.19215687f, 0.3647059f, 0.34117648f, 0.34901962f, 0.53333336f, 0.50980395f, 0.5176471f, 0.5647059f, 0.5411765f, 0.54901963f, 0.47058824f, 0.44705883f, 0.45490196f, 0.42745098f, 0.39215687f, 0.40392157f, 0.26666668f, 0.23137255f, 0.24313726f, 0.11372549f, 0.078431375f, 0.09019608f, 0.11372549f, 0.078431375f, 0.09019608f, 0.23921569f, 0.19607843f, 0.21176471f, 0.2627451f, 0.21960784f, 0.23529412f, 0.21568628f, 0.17254902f, 0.1882353f, 0.2509804f, 0.20784314f, 0.22352941f, 0.14509805f, 0.14509805f, 0.15294118f, 0.14901961f, 0.14901961f, 0.15686275f, 0.24705882f, 0.2509804f, 0.25882354f, 0.3019608f, 0.30588236f, 0.3137255f, 0.31764707f, 0.32156864f, 0.32941177f, 0.20784314f, 0.21176471f, 0.21960784f, 0.2f, 0.20392157f, 0.21176471f, 0.14901961f, 0.15294118f, 0.16078432f, 0.2f, 0.20392157f, 0.21176471f, 0.2784314f, 0.28235295f, 0.2901961f, 0.2509804f, 0.2509804f, 0.2509804f, 0.25490198f, 0.25490198f, 0.25490198f, 0.2901961f, 0.28235295f, 0.28627452f, 0.29803923f, 0.2901961f, 0.29411766f, 0.2627451f, 0.25490198f, 0.25882354f, 0.21960784f, 0.21176471f, 0.21568628f, 0.15294118f, 0.13333334f, 0.14901961f, 0.1882353f, 0.16862746f, 0.18431373f, 0.3647059f, 0.34509805f, 0.36078432f, 0.47058824f, 0.4509804f, 0.46666667f, 0.56078434f, 0.5372549f, 0.54509807f, 0.5647059f, 0.5411765f, 0.54901963f, 0.5372549f, 0.5137255f, 0.52156866f, 0.42745098f, 0.40392157f, 0.4117647f, 0.38431373f, 0.34901962f, 0.36078432f, 0.27450982f, 0.23921569f, 0.2509804f, 0.18039216f, 0.14509805f, 0.15686275f, 0.18431373f, 0.14901961f, 0.16078432f, 0.2627451f, 0.21960784f, 0.23529412f, 0.26666668f, 0.22352941f, 0.23921569f, 0.21176471f, 0.16862746f, 0.18431373f, 0.22745098f, 0.18431373f, 0.2f, 0.1882353f, 0.18431373f, 0.20392157f, 0.18039216f, 0.1764706f, 0.19607843f, 0.25882354f, 0.25882354f, 0.26666668f, 0.25490198f, 0.25490198f, 0.2627451f, 0.25490198f, 0.25882354f, 0.26666668f, 0.14509805f, 0.14901961f, 0.15686275f, 0.1882353f, 0.19215687f, 0.2f, 0.16078432f, 0.16470589f, 0.17254902f, 0.29411766f, 0.29803923f, 0.30588236f, 0.3647059f, 0.36862746f, 0.3764706f, 0.2784314f, 0.2784314f, 0.28627452f, 0.24705882f, 0.24705882f, 0.25490198f, 0.28235295f, 0.27450982f, 0.28627452f, 0.27058825f, 0.2627451f, 0.27450982f, 0.20784314f, 0.2f, 0.21176471f, 0.23137255f, 0.22352941f, 0.23529412f, 0.47843137f, 0.4627451f, 0.46666667f, 0.43137255f, 0.41568628f, 0.41960785f, 0.60784316f, 0.5921569f, 0.59607846f, 0.56078434f, 0.54509807f, 0.54901963f, 0.5019608f, 0.47843137f, 0.4862745f, 0.40784314f, 0.38431373f, 0.39215687f, 0.41960785f, 0.39607844f, 0.40392157f, 0.34509805f, 0.32156864f, 0.32941177f, 0.25490198f, 0.21960784f, 0.23137255f, 0.2509804f, 0.21568628f, 0.22745098f, 0.27450982f, 0.23921569f, 0.2509804f, 0.2784314f, 0.24313726f, 0.25490198f, 0.28627452f, 0.24705882f, 0.2509804f, 0.30588236f, 0.26666668f, 0.27058825f, 0.3529412f, 0.3137255f, 0.31764707f, 0.43137255f, 0.39215687f, 0.39607844f, 0.1764706f, 0.17254902f, 0.19215687f, 0.17254902f, 0.16862746f, 0.1882353f, 0.23921569f, 0.23921569f, 0.24705882f, 0.25882354f, 0.25882354f, 0.26666668f, 0.2509804f, 0.25490198f, 0.2627451f, 0.16862746f, 0.17254902f, 0.18039216f, 0.18431373f, 0.1882353f, 0.19607843f, 0.14509805f, 0.14901961f, 0.15686275f, 0.28627452f, 0.2901961f, 0.29803923f, 0.33333334f, 0.3372549f, 0.34509805f, 0.2627451f, 0.2627451f, 0.27058825f, 0.27058825f, 0.27058825f, 0.2784314f, 0.29803923f, 0.2901961f, 0.3019608f, 0.26666668f, 0.25882354f, 0.27058825f, 0.20392157f, 0.19607843f, 0.20784314f, 0.2784314f, 0.27058825f, 0.28235295f, 0.4392157f, 0.42352942f, 0.42745098f, 0.4509804f, 0.43529412f, 0.4392157f, 0.6f, 0.58431375f, 0.5882353f, 0.5568628f, 0.5411765f, 0.54509807f, 0.49803922f, 0.4745098f, 0.48235294f, 0.42352942f, 0.4f, 0.40784314f, 0.44313726f, 0.41960785f, 0.42745098f, 0.34117648f, 0.31764707f, 0.3254902f, 0.2627451f, 0.22745098f, 0.23921569f, 0.25490198f, 0.21960784f, 0.23137255f, 0.28627452f, 0.2509804f, 0.2627451f, 0.2784314f, 0.24313726f, 0.25490198f, 0.26666668f, 0.22745098f, 0.23137255f, 0.30980393f, 0.27058825f, 0.27450982f, 0.36078432f, 0.32156864f, 0.3254902f, 0.40392157f, 0.3647059f, 0.36862746f, 0.2784314f, 0.27450982f, 0.29411766f, 0.25882354f, 0.25490198f, 0.27450982f, 0.25882354f, 0.25490198f, 0.27450982f, 0.2784314f, 0.27450982f, 0.29411766f, 0.24313726f, 0.24705882f, 0.2627451f, 0.21568628f, 0.21960784f, 0.23529412f, 0.23529412f, 0.23921569f, 0.25490198f, 0.21568628f, 0.21960784f, 0.23529412f, 0.35686275f, 0.36078432f, 0.36862746f, 0.39607844f, 0.4f, 0.40784314f, 0.3764706f, 0.3764706f, 0.38431373f, 0.35686275f, 0.35686275f, 0.3647059f, 0.34509805f, 0.3372549f, 0.34901962f, 0.3647059f, 0.35686275f, 0.36862746f, 0.41960785f, 0.4f, 0.41568628f, 0.54509807f, 0.5254902f, 0.5411765f, 0.64705884f, 0.6313726f, 0.63529414f, 0.654902f, 0.6392157f, 0.6431373f, 0.59607846f, 0.5803922f, 0.58431375f, 0.49019608f, 0.4745098f, 0.47843137f, 0.39607844f, 0.37254903f, 0.38039216f, 0.30980393f, 0.28627452f, 0.29411766f, 0.32941177f, 0.30588236f, 0.3137255f, 0.24705882f, 0.22352941f, 0.23137255f, 0.21176471f, 0.1764706f, 0.18039216f, 0.20392157f, 0.16862746f, 0.17254902f, 0.2509804f, 0.21568628f, 0.21960784f, 0.25882354f, 0.22352941f, 0.22745098f, 0.26666668f, 0.22745098f, 0.23137255f, 0.3254902f, 0.28627452f, 0.2901961f, 0.3529412f, 0.3137255f, 0.31764707f, 0.34901962f, 0.30980393f, 0.3137255f, 0.3647059f, 0.36078432f, 0.38039216f, 0.3647059f, 0.36078432f, 0.38039216f, 0.3529412f, 0.34901962f, 0.36862746f, 0.3764706f, 0.37254903f, 0.39215687f, 0.3019608f, 0.30588236f, 0.32156864f, 0.27450982f, 0.2784314f, 0.29411766f, 0.25490198f, 0.25882354f, 0.27450982f, 0.22745098f, 0.23137255f, 0.24705882f, 0.32941177f, 0.33333334f, 0.34117648f, 0.34509805f, 0.34901962f, 0.35686275f, 0.35686275f, 0.35686275f, 0.3647059f, 0.32941177f, 0.32941177f, 0.3372549f, 0.40392157f, 0.39607844f, 0.40784314f, 0.5137255f, 0.5058824f, 0.5176471f, 0.58431375f, 0.5647059f, 0.5803922f, 0.6117647f, 0.5921569f, 0.60784316f, 0.54901963f, 0.53333336f, 0.5372549f, 0.5647059f, 0.54901963f, 0.5529412f, 0.41960785f, 0.40392157f, 0.40784314f, 0.3647059f, 0.34901962f, 0.3529412f, 0.27450982f, 0.2509804f, 0.25882354f, 0.16862746f, 0.14509805f, 0.15294118f, 0.1764706f, 0.15294118f, 0.16078432f, 0.14901961f, 0.1254902f, 0.13333334f, 0.2509804f, 0.21568628f, 0.21960784f, 0.24313726f, 0.20784314f, 0.21176471f, 0.27450982f, 0.23921569f, 0.24313726f, 0.26666668f, 0.23137255f, 0.23529412f, 0.26666668f, 0.22745098f, 0.23137255f, 0.30980393f, 0.27058825f, 0.27450982f, 0.32156864f, 0.28235295f, 0.28627452f, 0.30980393f, 0.27058825f, 0.27450982f, 0.34509805f, 0.34117648f, 0.3647059f, 0.33333334f, 0.32941177f, 0.3529412f, 0.32156864f, 0.31764707f, 0.3372549f, 0.3372549f, 0.33333334f, 0.3529412f, 0.2901961f, 0.29411766f, 0.30980393f, 0.2627451f, 0.26666668f, 0.28235295f, 0.25882354f, 0.2627451f, 0.2784314f, 0.21960784f, 0.22352941f, 0.23921569f, 0.2f, 0.2f, 0.20784314f, 0.23921569f, 0.23921569f, 0.24705882f, 0.31764707f, 0.31764707f, 0.3254902f, 0.3529412f, 0.3529412f, 0.36078432f, 0.53333336f, 0.5254902f, 0.5372549f, 0.60784316f, 0.6f, 0.6117647f, 0.5294118f, 0.50980395f, 0.5254902f, 0.4f, 0.38039216f, 0.39607844f, 0.36862746f, 0.3529412f, 0.35686275f, 0.38431373f, 0.36862746f, 0.37254903f, 0.30980393f, 0.29411766f, 0.29803923f, 0.32156864f, 0.30588236f, 0.30980393f, 0.26666668f, 0.24313726f, 0.2509804f, 0.18431373f, 0.16078432f, 0.16862746f, 0.18039216f, 0.15686275f, 0.15686275f, 0.1764706f, 0.15294118f, 0.15294118f, 0.28627452f, 0.2509804f, 0.25490198f, 0.23921569f, 0.20392157f, 0.20784314f, 0.22352941f, 0.1882353f, 0.19215687f, 0.22352941f, 0.1882353f, 0.19215687f, 0.27450982f, 0.23529412f, 0.23137255f, 0.34117648f, 0.3019608f, 0.29803923f, 0.32941177f, 0.2901961f, 0.28627452f, 0.29411766f, 0.25490198f, 0.2509804f, 0.18431373f, 0.18039216f, 0.20392157f, 0.19607843f, 0.19215687f, 0.21568628f, 0.2627451f, 0.25882354f, 0.2784314f, 0.3254902f, 0.32156864f, 0.34117648f, 0.34901962f, 0.3529412f, 0.36862746f, 0.3019608f, 0.30588236f, 0.32156864f, 0.25882354f, 0.2627451f, 0.2784314f, 0.16470589f, 0.16862746f, 0.1764706f, 0.1882353f, 0.1882353f, 0.19607843f, 0.34901962f, 0.34901962f, 0.35686275f, 0.52156866f, 0.52156866f, 0.5294118f, 0.49019608f, 0.49019608f, 0.49803922f, 0.5411765f, 0.53333336f, 0.54509807f, 0.45490196f, 0.44705883f, 0.45882353f, 0.3647059f, 0.34509805f, 0.36078432f, 0.3137255f, 0.29803923f, 0.3019608f, 0.3254902f, 0.30980393f, 0.3137255f, 0.3137255f, 0.29803923f, 0.3019608f, 0.2901961f, 0.27450982f, 0.2784314f, 0.29803923f, 0.28235295f, 0.28627452f, 0.2784314f, 0.25490198f, 0.2627451f, 0.27058825f, 0.24705882f, 0.25490198f, 0.2901961f, 0.26666668f, 0.26666668f, 0.29411766f, 0.27058825f, 0.27058825f, 0.2509804f, 0.21568628f, 0.21960784f, 0.23921569f, 0.20392157f, 0.20784314f, 0.24705882f, 0.21176471f, 0.21568628f, 0.23921569f, 0.20392157f, 0.20784314f, 0.27058825f, 0.23137255f, 0.22745098f, 0.32156864f, 0.28235295f, 0.2784314f, 0.3254902f, 0.28627452f, 0.28235295f, 0.3254902f, 0.28627452f, 0.28235295f, 0.1882353f, 0.18431373f, 0.20392157f, 0.19215687f, 0.1882353f, 0.20784314f, 0.27450982f, 0.27058825f, 0.2901961f, 0.3254902f, 0.32156864f, 0.34117648f, 0.3372549f, 0.34117648f, 0.34901962f, 0.29411766f, 0.29803923f, 0.30588236f, 0.3019608f, 0.30588236f, 0.3137255f, 0.2784314f, 0.28235295f, 0.2901961f, 0.34509805f, 0.34901962f, 0.35686275f, 0.4862745f, 0.49411765f, 0.49019608f, 0.42745098f, 0.43529412f, 0.43137255f, 0.4392157f, 0.44705883f, 0.44313726f, 0.39607844f, 0.39607844f, 0.39607844f, 0.3647059f, 0.3647059f, 0.3647059f, 0.30980393f, 0.3019608f, 0.30588236f, 0.2901961f, 0.28235295f, 0.28627452f, 0.28235295f, 0.27450982f, 0.2784314f, 0.2627451f, 0.25490198f, 0.26666668f, 0.3019608f, 0.28235295f, 0.29803923f, 0.30980393f, 0.2901961f, 0.30588236f, 0.2784314f, 0.2627451f, 0.26666668f, 0.32156864f, 0.30588236f, 0.30980393f, 0.3254902f, 0.3019608f, 0.30980393f, 0.34901962f, 0.3254902f, 0.33333334f, 0.27450982f, 0.23921569f, 0.2509804f, 0.23921569f, 0.20392157f, 0.21568628f, 0.23529412f, 0.2f, 0.20392157f, 0.23137255f, 0.19607843f, 0.2f, 0.27450982f, 0.23529412f, 0.23921569f, 0.33333334f, 0.29411766f, 0.29803923f, 0.30980393f, 0.27058825f, 0.27450982f, 0.33333334f, 0.29411766f, 0.29803923f, 0.19215687f, 0.1882353f, 0.20784314f, 0.2f, 0.19607843f, 0.21568628f, 0.28235295f, 0.2784314f, 0.29803923f, 0.34117648f, 0.3372549f, 0.35686275f, 0.3372549f, 0.34117648f, 0.34901962f, 0.2901961f, 0.29411766f, 0.3019608f, 0.28235295f, 0.28627452f, 0.29411766f, 0.30588236f, 0.30980393f, 0.31764707f, 0.38431373f, 0.39215687f, 0.3882353f, 0.47843137f, 0.4862745f, 0.48235294f, 0.43529412f, 0.44313726f, 0.4392157f, 0.44313726f, 0.4509804f, 0.44705883f, 0.40784314f, 0.40784314f, 0.40784314f, 0.35686275f, 0.35686275f, 0.35686275f, 0.29411766f, 0.2901961f, 0.28235295f, 0.28235295f, 0.2784314f, 0.27058825f, 0.28235295f, 0.27450982f, 0.28627452f, 0.26666668f, 0.25882354f, 0.27058825f, 0.2784314f, 0.25882354f, 0.27450982f, 0.28235295f, 0.2627451f, 0.2784314f, 0.2627451f, 0.24705882f, 0.2509804f, 0.30588236f, 0.2901961f, 0.29411766f, 0.3254902f, 0.3019608f, 0.30980393f, 0.38039216f, 0.35686275f, 0.3647059f, 0.2784314f, 0.24313726f, 0.25490198f, 0.21960784f, 0.18431373f, 0.19607843f, 0.23137255f, 0.19607843f, 0.2f, 0.25882354f, 0.22352941f, 0.22745098f, 0.30588236f, 0.26666668f, 0.27058825f, 0.32941177f, 0.2901961f, 0.29411766f, 0.29803923f, 0.25882354f, 0.2627451f, 0.31764707f, 0.2784314f, 0.28235295f, 0.25882354f, 0.25490198f, 0.27450982f, 0.25882354f, 0.25490198f, 0.27450982f, 0.30980393f, 0.30588236f, 0.3254902f, 0.3254902f, 0.32156864f, 0.34117648f, 0.29411766f, 0.29803923f, 0.3137255f, 0.27058825f, 0.27450982f, 0.2901961f, 0.2627451f, 0.26666668f, 0.27450982f, 0.34901962f, 0.3529412f, 0.36078432f, 0.42745098f, 0.43137255f, 0.4392157f, 0.38039216f, 0.38431373f, 0.39215687f, 0.30588236f, 0.3137255f, 0.30980393f, 0.3137255f, 0.32156864f, 0.31764707f, 0.35686275f, 0.35686275f, 0.35686275f, 0.3254902f, 0.3254902f, 0.3254902f, 0.26666668f, 0.2627451f, 0.25490198f, 0.25882354f, 0.25490198f, 0.24705882f, 0.28235295f, 0.27450982f, 0.28627452f, 0.29803923f, 0.2901961f, 0.3019608f, 0.3137255f, 0.29411766f, 0.30980393f, 0.32941177f, 0.30980393f, 0.3254902f, 0.3254902f, 0.30980393f, 0.3137255f, 0.33333334f, 0.31764707f, 0.32156864f, 0.31764707f, 0.29411766f, 0.3019608f, 0.37254903f, 0.34901962f, 0.35686275f, 0.34117648f, 0.30588236f, 0.31764707f, 0.21568628f, 0.18039216f, 0.19215687f, 0.20784314f, 0.17254902f, 0.18431373f, 0.2509804f, 0.21568628f, 0.22745098f, 0.30980393f, 0.27058825f, 0.27450982f, 0.32156864f, 0.28235295f, 0.28627452f, 0.2901961f, 0.2509804f, 0.25490198f, 0.3137255f, 0.27450982f, 0.2784314f, 0.29411766f, 0.2901961f, 0.30980393f, 0.30980393f, 0.30588236f, 0.3254902f, 0.3137255f, 0.30980393f, 0.32941177f, 0.2901961f, 0.28627452f, 0.30588236f, 0.2627451f, 0.26666668f, 0.28235295f, 0.30980393f, 0.3137255f, 0.32941177f, 0.30588236f, 0.30980393f, 0.31764707f, 0.39607844f, 0.4f, 0.40784314f, 0.4117647f, 0.41568628f, 0.42352942f, 0.36862746f, 0.37254903f, 0.38039216f, 0.3882353f, 0.39607844f, 0.39215687f, 0.39607844f, 0.40392157f, 0.4f, 0.39215687f, 0.39215687f, 0.39215687f, 0.28627452f, 0.28627452f, 0.28627452f, 0.23137255f, 0.22745098f, 0.21960784f, 0.2509804f, 0.24313726f, 0.24705882f, 0.29411766f, 0.28627452f, 0.29803923f, 0.3019608f, 0.29411766f, 0.30588236f, 0.2784314f, 0.25882354f, 0.27450982f, 0.29803923f, 0.2784314f, 0.29411766f, 0.32156864f, 0.30588236f, 0.30980393f, 0.32941177f, 0.3137255f, 0.31764707f, 0.3254902f, 0.3019608f, 0.30980393f, 0.40392157f, 0.38039216f, 0.3882353f, 0.43137255f, 0.39607844f, 0.40784314f, 0.25882354f, 0.22352941f, 0.23529412f, 0.21960784f, 0.18431373f, 0.19607843f, 0.24313726f, 0.20784314f, 0.21960784f, 0.31764707f, 0.2784314f, 0.28235295f, 0.32156864f, 0.28235295f, 0.28627452f, 0.2901961f, 0.2509804f, 0.25490198f, 0.29803923f, 0.25882354f, 0.2627451f, 0.31764707f, 0.30980393f, 0.32941177f, 0.3137255f, 0.30588236f, 0.3254902f, 0.2509804f, 0.24705882f, 0.26666668f, 0.20784314f, 0.20392157f, 0.22352941f, 0.25490198f, 0.2509804f, 0.27058825f, 0.3882353f, 0.38431373f, 0.40392157f, 0.40784314f, 0.4117647f, 0.41960785f, 0.44313726f, 0.44705883f, 0.45490196f, 0.36862746f, 0.37254903f, 0.38039216f, 0.34901962f, 0.3529412f, 0.36078432f, 0.43529412f, 0.44313726f, 0.4392157f, 0.44313726f, 0.4509804f, 0.44705883f, 0.39215687f, 0.39215687f, 0.39215687f, 0.2627451f, 0.2627451f, 0.2627451f, 0.2509804f, 0.24313726f, 0.24705882f, 0.29411766f, 0.28627452f, 0.2901961f, 0.31764707f, 0.30980393f, 0.32156864f, 0.3137255f, 0.30588236f, 0.31764707f, 0.25882354f, 0.23921569f, 0.25490198f, 0.29411766f, 0.27450982f, 0.2901961f, 0.3529412f, 0.33333334f, 0.34901962f, 0.3647059f, 0.34509805f, 0.36078432f, 0.3764706f, 0.3529412f, 0.36078432f, 0.46666667f, 0.44313726f, 0.4509804f, 0.47843137f, 0.44313726f, 0.45490196f, 0.32941177f, 0.29411766f, 0.30588236f, 0.2901961f, 0.25490198f, 0.26666668f, 0.2901961f, 0.25490198f, 0.26666668f, 0.36078432f, 0.31764707f, 0.33333334f, 0.3529412f, 0.30980393f, 0.3254902f, 0.31764707f, 0.27450982f, 0.2901961f, 0.32156864f, 0.2784314f, 0.29411766f, 0.27058825f, 0.2627451f, 0.28235295f, 0.26666668f, 0.25882354f, 0.2784314f, 0.22352941f, 0.21960784f, 0.23921569f, 0.23529412f, 0.23137255f, 0.2509804f, 0.32941177f, 0.3254902f, 0.34509805f, 0.43137255f, 0.42745098f, 0.44705883f, 0.40784314f, 0.4117647f, 0.41960785f, 0.3372549f, 0.34117648f, 0.34901962f, 0.3372549f, 0.34117648f, 0.34901962f, 0.29803923f, 0.3019608f, 0.30980393f, 0.33333334f, 0.34117648f, 0.3372549f, 0.3254902f, 0.33333334f, 0.32941177f, 0.2784314f, 0.2784314f, 0.2784314f, 0.22352941f, 0.22352941f, 0.22352941f, 0.30588236f, 0.29803923f, 0.3019608f, 0.3647059f, 0.35686275f, 0.36078432f, 0.30588236f, 0.29803923f, 0.30980393f, 0.2901961f, 0.28235295f, 0.29411766f, 0.23137255f, 0.21176471f, 0.22745098f, 0.28627452f, 0.26666668f, 0.28235295f, 0.34117648f, 0.32156864f, 0.3372549f, 0.33333334f, 0.3137255f, 0.32941177f, 0.33333334f, 0.30980393f, 0.31764707f, 0.40392157f, 0.38039216f, 0.3882353f, 0.4627451f, 0.42745098f, 0.4392157f, 0.36078432f, 0.3254902f, 0.3372549f, 0.34509805f, 0.30980393f, 0.32156864f, 0.3254902f, 0.2901961f, 0.3019608f, 0.40784314f, 0.3647059f, 0.38039216f, 0.39607844f, 0.3529412f, 0.36862746f, 0.3764706f, 0.33333334f, 0.34901962f, 0.4117647f, 0.36862746f, 0.38431373f, 0.29803923f, 0.28627452f, 0.3137255f, 0.27058825f, 0.25882354f, 0.28627452f, 0.22352941f, 0.21568628f, 0.23529412f, 0.2627451f, 0.25490198f, 0.27450982f, 0.33333334f, 0.32941177f, 0.34901962f, 0.36862746f, 0.3647059f, 0.38431373f, 0.40784314f, 0.4117647f, 0.42745098f, 0.35686275f, 0.36078432f, 0.3764706f, 0.3254902f, 0.32941177f, 0.3372549f, 0.3254902f, 0.32941177f, 0.3372549f, 0.34117648f, 0.34117648f, 0.34117648f, 0.32156864f, 0.32156864f, 0.32156864f, 0.23137255f, 0.23137255f, 0.23137255f, 0.2f, 0.2f, 0.2f, 0.32941177f, 0.32941177f, 0.32941177f, 0.37254903f, 0.37254903f, 0.37254903f, 0.31764707f, 0.30980393f, 0.32156864f, 0.2901961f, 0.28235295f, 0.29411766f, 0.22352941f, 0.20392157f, 0.21960784f, 0.28627452f, 0.26666668f, 0.28235295f, 0.33333334f, 0.3137255f, 0.32941177f, 0.31764707f, 0.29803923f, 0.3137255f, 0.32941177f, 0.30588236f, 0.32156864f, 0.39607844f, 0.37254903f, 0.3882353f, 0.4392157f, 0.40392157f, 0.41568628f, 0.34901962f, 0.3137255f, 0.3254902f, 0.3372549f, 0.3019608f, 0.3137255f, 0.3254902f, 0.2901961f, 0.3019608f, 0.44705883f, 0.40392157f, 0.41960785f, 0.41568628f, 0.37254903f, 0.3882353f, 0.38039216f, 0.3372549f, 0.3529412f, 0.42745098f, 0.38431373f, 0.4f, 0.19215687f, 0.18039216f, 0.20784314f, 0.22352941f, 0.21176471f, 0.23921569f, 0.2901961f, 0.28235295f, 0.3019608f, 0.41568628f, 0.40784314f, 0.42745098f, 0.4509804f, 0.44705883f, 0.46666667f, 0.3647059f, 0.36078432f, 0.38039216f, 0.40392157f, 0.4f, 0.41960785f, 0.3529412f, 0.35686275f, 0.37254903f, 0.31764707f, 0.32156864f, 0.32941177f, 0.29411766f, 0.29803923f, 0.30588236f, 0.2627451f, 0.2627451f, 0.2627451f, 0.27058825f, 0.27058825f, 0.27058825f, 0.22745098f, 0.22745098f, 0.22745098f, 0.24313726f, 0.24313726f, 0.24313726f, 0.3529412f, 0.3529412f, 0.3529412f, 0.3254902f, 0.3254902f, 0.3254902f, 0.20392157f, 0.19607843f, 0.20784314f, 0.21176471f, 0.20392157f, 0.21568628f, 0.20784314f, 0.1882353f, 0.20392157f, 0.32156864f, 0.3019608f, 0.31764707f, 0.37254903f, 0.3529412f, 0.36862746f, 0.34117648f, 0.32156864f, 0.3372549f, 0.33333334f, 0.30980393f, 0.3254902f, 0.3764706f, 0.3529412f, 0.36862746f, 0.42352942f, 0.3882353f, 0.4f, 0.32941177f, 0.29411766f, 0.30588236f, 0.31764707f, 0.28235295f, 0.29411766f, 0.33333334f, 0.29803923f, 0.30980393f, 0.4862745f, 0.44313726f, 0.45882353f, 0.41568628f, 0.37254903f, 0.3882353f, 0.31764707f, 0.27450982f, 0.2901961f, 0.33333334f, 0.2901961f, 0.30588236f, 0.19607843f, 0.1764706f, 0.20392157f, 0.18039216f, 0.16078432f, 0.1882353f, 0.34117648f, 0.32941177f, 0.35686275f, 0.44313726f, 0.43137255f, 0.45882353f, 0.4392157f, 0.43137255f, 0.4509804f, 0.3529412f, 0.34901962f, 0.36862746f, 0.33333334f, 0.32941177f, 0.34901962f, 0.3137255f, 0.30980393f, 0.32941177f, 0.27450982f, 0.2784314f, 0.28627452f, 0.26666668f, 0.27058825f, 0.2784314f, 0.2509804f, 0.2509804f, 0.25882354f, 0.24705882f, 0.24705882f, 0.25490198f, 0.3137255f, 0.3137255f, 0.3137255f, 0.36078432f, 0.36078432f, 0.36078432f, 0.34509805f, 0.34509805f, 0.34509805f, 0.29803923f, 0.29803923f, 0.29803923f, 0.2627451f, 0.25490198f, 0.26666668f, 0.25490198f, 0.24705882f, 0.25882354f, 0.32941177f, 0.30980393f, 0.3254902f, 0.38039216f, 0.36078432f, 0.3764706f, 0.44705883f, 0.42745098f, 0.44313726f, 0.38431373f, 0.3647059f, 0.38039216f, 0.35686275f, 0.33333334f, 0.34901962f, 0.3882353f, 0.3647059f, 0.38039216f, 0.39607844f, 0.36078432f, 0.38039216f, 0.22352941f, 0.1882353f, 0.20784314f, 0.29411766f, 0.25882354f, 0.2784314f, 0.29803923f, 0.2627451f, 0.28235295f, 0.46666667f, 0.42352942f, 0.44705883f, 0.47058824f, 0.42745098f, 0.4509804f, 0.2509804f, 0.20784314f, 0.23137255f, 0.25490198f, 0.21176471f, 0.23529412f, 0.2901961f, 0.27058825f, 0.29803923f, 0.28235295f, 0.2627451f, 0.2901961f, 0.3882353f, 0.36862746f, 0.39607844f, 0.3882353f, 0.3764706f, 0.40392157f, 0.33333334f, 0.3254902f, 0.34509805f, 0.28235295f, 0.27450982f, 0.29411766f, 0.29803923f, 0.29411766f, 0.3137255f, 0.29803923f, 0.29411766f, 0.3137255f, 0.2627451f, 0.26666668f, 0.27450982f, 0.25490198f, 0.25882354f, 0.26666668f, 0.25490198f, 0.25490198f, 0.2627451f, 0.25490198f, 0.25490198f, 0.2627451f, 0.27058825f, 0.27058825f, 0.27058825f, 0.28627452f, 0.28627452f, 0.28627452f, 0.29411766f, 0.29411766f, 0.29411766f, 0.28627452f, 0.28627452f, 0.28627452f, 0.29411766f, 0.28627452f, 0.29803923f, 0.28627452f, 0.2784314f, 0.2901961f, 0.30980393f, 0.2901961f, 0.30588236f, 0.34509805f, 0.3254902f, 0.34117648f, 0.34509805f, 0.3254902f, 0.34117648f, 0.36078432f, 0.34117648f, 0.35686275f, 0.3647059f, 0.34117648f, 0.35686275f, 0.42745098f, 0.40392157f, 0.41960785f, 0.3647059f, 0.32941177f, 0.34901962f, 0.21568628f, 0.18039216f, 0.2f, 0.3137255f, 0.2784314f, 0.29803923f, 0.3137255f, 0.2784314f, 0.29803923f, 0.47843137f, 0.43529412f, 0.45882353f, 0.5019608f, 0.45882353f, 0.48235294f, 0.35686275f, 0.3137255f, 0.3372549f, 0.36862746f, 0.3254902f, 0.34901962f, 0.4117647f, 0.38431373f, 0.41568628f, 0.40784314f, 0.3882353f, 0.41568628f, 0.44705883f, 0.42745098f, 0.45490196f, 0.3529412f, 0.33333334f, 0.36078432f, 0.21568628f, 0.20392157f, 0.23137255f, 0.16862746f, 0.15686275f, 0.18431373f, 0.18431373f, 0.18039216f, 0.2f, 0.18039216f, 0.1764706f, 0.19607843f, 0.25882354f, 0.25490198f, 0.27450982f, 0.25882354f, 0.25490198f, 0.27450982f, 0.28627452f, 0.28627452f, 0.29411766f, 0.29411766f, 0.29411766f, 0.3019608f, 0.24313726f, 0.24313726f, 0.2509804f, 0.22745098f, 0.22745098f, 0.23529412f, 0.25882354f, 0.25882354f, 0.25882354f, 0.28235295f, 0.28235295f, 0.28235295f, 0.29803923f, 0.2901961f, 0.3019608f, 0.30980393f, 0.3019608f, 0.3137255f, 0.3254902f, 0.30588236f, 0.32156864f, 0.3764706f, 0.35686275f, 0.37254903f, 0.32156864f, 0.3019608f, 0.31764707f, 0.3882353f, 0.36862746f, 0.38431373f, 0.40392157f, 0.38039216f, 0.39607844f, 0.48235294f, 0.45882353f, 0.4745098f, 0.4f, 0.3647059f, 0.38431373f, 0.22745098f, 0.19215687f, 0.21176471f, 0.3137255f, 0.2784314f, 0.29803923f, 0.3019608f, 0.26666668f, 0.28627452f, 0.47058824f, 0.42745098f, 0.45882353f, 0.49803922f, 0.45490196f, 0.4862745f, 0.41568628f, 0.37254903f, 0.40392157f, 0.39607844f, 0.3529412f, 0.38431373f, 0.42352942f, 0.39607844f, 0.42745098f, 0.43137255f, 0.40392157f, 0.43529412f, 0.44313726f, 0.42352942f, 0.4509804f, 0.33333334f, 0.3137255f, 0.34117648f, 0.2f, 0.1882353f, 0.21568628f, 0.18431373f, 0.17254902f, 0.2f, 0.20392157f, 0.2f, 0.21960784f, 0.20392157f, 0.2f, 0.21960784f, 0.23529412f, 0.23137255f, 0.2509804f, 0.25490198f, 0.2509804f, 0.27058825f, 0.29803923f, 0.29803923f, 0.30588236f, 0.30588236f, 0.30588236f, 0.3137255f, 0.21960784f, 0.21960784f, 0.22745098f, 0.21176471f, 0.21176471f, 0.21960784f, 0.24705882f, 0.24705882f, 0.24705882f, 0.27058825f, 0.27058825f, 0.27058825f, 0.29803923f, 0.2901961f, 0.3019608f, 0.30588236f, 0.29803923f, 0.30980393f, 0.3254902f, 0.30588236f, 0.32156864f, 0.3764706f, 0.35686275f, 0.37254903f, 0.3019608f, 0.28235295f, 0.29803923f, 0.36862746f, 0.34901962f, 0.3647059f, 0.39215687f, 0.36862746f, 0.38431373f, 0.46666667f, 0.44313726f, 0.45882353f, 0.41568628f, 0.38039216f, 0.4f, 0.23921569f, 0.20392157f, 0.22352941f, 0.3254902f, 0.2901961f, 0.30980393f, 0.32156864f, 0.28627452f, 0.30588236f, 0.47058824f, 0.42745098f, 0.45882353f, 0.4627451f, 0.41960785f, 0.4509804f, 0.42745098f, 0.38431373f, 0.41568628f, 0.38431373f, 0.34117648f, 0.37254903f, 0.41568628f, 0.3882353f, 0.41960785f, 0.39607844f, 0.36862746f, 0.4f, 0.36862746f, 0.34901962f, 0.3764706f, 0.2784314f, 0.25882354f, 0.28627452f, 0.17254902f, 0.16078432f, 0.1882353f, 0.16078432f, 0.14901961f, 0.1764706f, 0.15686275f, 0.15294118f, 0.17254902f, 0.15686275f, 0.15294118f, 0.17254902f, 0.23137255f, 0.22745098f, 0.24705882f, 0.2509804f, 0.24705882f, 0.26666668f, 0.27450982f, 0.27450982f, 0.28235295f, 0.27450982f, 0.27450982f, 0.28235295f, 0.20392157f, 0.20392157f, 0.21176471f, 0.24313726f, 0.24313726f, 0.2509804f, 0.27450982f, 0.27450982f, 0.28235295f, 0.27450982f, 0.27450982f, 0.28235295f, 0.3529412f, 0.34509805f, 0.34901962f, 0.3529412f, 0.34509805f, 0.34901962f, 0.41960785f, 0.4f, 0.41568628f, 0.4745098f, 0.45490196f, 0.47058824f, 0.41960785f, 0.4f, 0.41568628f, 0.4117647f, 0.39215687f, 0.40784314f, 0.4117647f, 0.3882353f, 0.40392157f, 0.45490196f, 0.43137255f, 0.44705883f, 0.39607844f, 0.36078432f, 0.38039216f, 0.25490198f, 0.21960784f, 0.23921569f, 0.34509805f, 0.30588236f, 0.3372549f, 0.36078432f, 0.32156864f, 0.3529412f, 0.47058824f, 0.42745098f, 0.45882353f, 0.43529412f, 0.39215687f, 0.42352942f, 0.47058824f, 0.42745098f, 0.45882353f, 0.45490196f, 0.4117647f, 0.44313726f, 0.32156864f, 0.29411766f, 0.3254902f, 0.3137255f, 0.28627452f, 0.31764707f, 0.32156864f, 0.3019608f, 0.32941177f, 0.33333334f, 0.3137255f, 0.34117648f, 0.3254902f, 0.30588236f, 0.33333334f, 0.3529412f, 0.34117648f, 0.36862746f, 0.34117648f, 0.3372549f, 0.35686275f, 0.34509805f, 0.34117648f, 0.36078432f, 0.2784314f, 0.27450982f, 0.29411766f, 0.2784314f, 0.27450982f, 0.29411766f, 0.24313726f, 0.24313726f, 0.2509804f, 0.24313726f, 0.24313726f, 0.2509804f, 0.22745098f, 0.22745098f, 0.23529412f, 0.3254902f, 0.3254902f, 0.33333334f, 0.34117648f, 0.34117648f, 0.34901962f, 0.30980393f, 0.30980393f, 0.31764707f, 0.3254902f, 0.31764707f, 0.32156864f, 0.3254902f, 0.31764707f, 0.32156864f, 0.4f, 0.38039216f, 0.39607844f, 0.44705883f, 0.42745098f, 0.44313726f, 0.43137255f, 0.4117647f, 0.42745098f, 0.40784314f, 0.3882353f, 0.40392157f, 0.40784314f, 0.38431373f, 0.4f, 0.4509804f, 0.42745098f, 0.44313726f, 0.4627451f, 0.42745098f, 0.44705883f, 0.30980393f, 0.27450982f, 0.29411766f, 0.3137255f, 0.27450982f, 0.30588236f, 0.32156864f, 0.28235295f, 0.3137255f, 0.43137255f, 0.3882353f, 0.41960785f, 0.43137255f, 0.3882353f, 0.41960785f, 0.5176471f, 0.4745098f, 0.5058824f, 0.50980395f, 0.46666667f, 0.49803922f, 0.16470589f, 0.13725491f, 0.16862746f, 0.17254902f, 0.14509805f, 0.1764706f, 0.21568628f, 0.1882353f, 0.21960784f, 0.28235295f, 0.2627451f, 0.2901961f, 0.32941177f, 0.30980393f, 0.3372549f, 0.39215687f, 0.38039216f, 0.40784314f, 0.37254903f, 0.36862746f, 0.3882353f, 0.3764706f, 0.37254903f, 0.39215687f, 0.29803923f, 0.29411766f, 0.3137255f, 0.27058825f, 0.26666668f, 0.28627452f, 0.18431373f, 0.18431373f, 0.19215687f, 0.2f, 0.2f, 0.20784314f, 0.24705882f, 0.24705882f, 0.25490198f, 0.3764706f, 0.3764706f, 0.38431373f, 0.35686275f, 0.35686275f, 0.3647059f, 0.30980393f, 0.30980393f, 0.31764707f, 0.3254902f, 0.31764707f, 0.32156864f, 0.3137255f, 0.30588236f, 0.30980393f, 0.33333334f, 0.31764707f, 0.32156864f, 0.32941177f, 0.3137255f, 0.31764707f, 0.32941177f, 0.30980393f, 0.3254902f, 0.34509805f, 0.3254902f, 0.34117648f, 0.38431373f, 0.36078432f, 0.3764706f, 0.4627451f, 0.4392157f, 0.45490196f, 0.49803922f, 0.45882353f, 0.49019608f, 0.34901962f, 0.30980393f, 0.34117648f, 0.27058825f, 0.23137255f, 0.2627451f, 0.27058825f, 0.23137255f, 0.2627451f, 0.39607844f, 0.3529412f, 0.38431373f, 0.43529412f, 0.39215687f, 0.42352942f, 0.5176471f, 0.4745098f, 0.5058824f, 0.4745098f, 0.43137255f, 0.4627451f, 0.18431373f, 0.15686275f, 0.1882353f, 0.18431373f, 0.15686275f, 0.1882353f, 0.19607843f, 0.16862746f, 0.2f, 0.24705882f, 0.21960784f, 0.2509804f, 0.28235295f, 0.2627451f, 0.2901961f, 0.3529412f, 0.34117648f, 0.36862746f, 0.3372549f, 0.32941177f, 0.34901962f, 0.34117648f, 0.3372549f, 0.35686275f, 0.25490198f, 0.2509804f, 0.27058825f, 0.20784314f, 0.20392157f, 0.22352941f, 0.101960786f, 0.101960786f, 0.10980392f, 0.13725491f, 0.13725491f, 0.14509805f, 0.21960784f, 0.21960784f, 0.22745098f, 0.3529412f, 0.3529412f, 0.36078432f, 0.30588236f, 0.30588236f, 0.3137255f, 0.24705882f, 0.24705882f, 0.25490198f, 0.3019608f, 0.29411766f, 0.29803923f, 0.33333334f, 0.3254902f, 0.32941177f, 0.34509805f, 0.32941177f, 0.33333334f, 0.33333334f, 0.31764707f, 0.32156864f, 0.30980393f, 0.2901961f, 0.30588236f, 0.3137255f, 0.29411766f, 0.30980393f, 0.3019608f, 0.2784314f, 0.29411766f, 0.35686275f, 0.33333334f, 0.34901962f, 0.4117647f, 0.37254903f, 0.40392157f, 0.3372549f, 0.29803923f, 0.32941177f, 0.27450982f, 0.23529412f, 0.26666668f, 0.2901961f, 0.2509804f, 0.28235295f, 0.40392157f, 0.36078432f, 0.39215687f, 0.44313726f, 0.4f, 0.43137255f, 0.49803922f, 0.45490196f, 0.4862745f, 0.41960785f, 0.3764706f, 0.40784314f,\n",
      "};\n",
      "\n",
      "cactus_X_test2_data.h\n",
      "\n",
      " X_test2[]:\n",
      " {\n",
      " 0.36862746f, 0.36078432f, 0.2784314f, 0.46666667f, 0.45882353f, 0.3764706f, 0.54901963f, 0.54509807f, 0.4627451f, 0.5411765f, 0.5372549f, 0.45490196f, 0.46666667f, 0.4745098f, 0.3882353f, 0.41960785f, 0.42745098f, 0.34117648f, 0.40784314f, 0.42352942f, 0.33333334f, 0.41960785f, 0.43529412f, 0.34509805f, 0.3647059f, 0.38431373f, 0.29411766f, 0.40392157f, 0.42352942f, 0.33333334f, 0.45882353f, 0.49019608f, 0.39607844f, 0.47058824f, 0.5019608f, 0.40784314f, 0.4392157f, 0.47843137f, 0.38039216f, 0.4627451f, 0.5019608f, 0.40392157f, 0.45490196f, 0.5058824f, 0.40392157f, 0.39607844f, 0.43529412f, 0.3372549f, 0.3882353f, 0.42352942f, 0.31764707f, 0.4117647f, 0.43529412f, 0.33333334f, 0.38431373f, 0.40784314f, 0.30588236f, 0.3647059f, 0.3882353f, 0.28627452f, 0.44313726f, 0.46666667f, 0.3647059f, 0.5647059f, 0.5882353f, 0.4862745f, 0.5764706f, 0.6f, 0.49803922f, 0.50980395f, 0.53333336f, 0.43137255f, 0.48235294f, 0.5058824f, 0.40392157f, 0.4f, 0.42352942f, 0.32156864f, 0.50980395f, 0.53333336f, 0.43137255f, 0.5019608f, 0.5254902f, 0.42352942f, 0.44313726f, 0.46666667f, 0.3647059f, 0.54901963f, 0.57254905f, 0.47058824f, 0.5058824f, 0.5294118f, 0.42745098f, 0.31764707f, 0.34117648f, 0.23921569f, 0.44313726f, 0.43529412f, 0.3529412f, 0.4627451f, 0.45490196f, 0.37254903f, 0.4745098f, 0.47058824f, 0.3882353f, 0.4745098f, 0.47058824f, 0.3882353f, 0.47058824f, 0.47843137f, 0.39215687f, 0.48235294f, 0.49019608f, 0.40392157f, 0.47843137f, 0.49411765f, 0.40392157f, 0.4745098f, 0.49019608f, 0.4f, 0.42352942f, 0.44313726f, 0.3529412f, 0.4509804f, 0.47058824f, 0.38039216f, 0.5058824f, 0.5372549f, 0.44313726f, 0.5176471f, 0.54901963f, 0.45490196f, 0.49019608f, 0.5294118f, 0.43137255f, 0.49803922f, 0.5372549f, 0.4392157f, 0.4627451f, 0.5137255f, 0.4117647f, 0.38431373f, 0.42352942f, 0.3254902f, 0.36862746f, 0.40392157f, 0.29803923f, 0.42352942f, 0.44705883f, 0.34509805f, 0.43529412f, 0.45882353f, 0.35686275f, 0.41568628f, 0.4392157f, 0.3372549f, 0.44313726f, 0.46666667f, 0.3647059f, 0.49411765f, 0.5176471f, 0.41568628f, 0.4745098f, 0.49803922f, 0.39607844f, 0.40784314f, 0.43137255f, 0.32941177f, 0.38039216f, 0.40392157f, 0.3019608f, 0.41568628f, 0.4392157f, 0.3372549f, 0.49019608f, 0.5137255f, 0.4117647f, 0.44705883f, 0.47058824f, 0.36862746f, 0.4509804f, 0.4745098f, 0.37254903f, 0.5254902f, 0.54901963f, 0.44705883f, 0.47058824f, 0.49411765f, 0.39215687f, 0.41568628f, 0.4392157f, 0.3372549f, 0.45882353f, 0.4509804f, 0.36862746f, 0.46666667f, 0.45882353f, 0.3764706f, 0.4745098f, 0.47058824f, 0.3882353f, 0.4862745f, 0.48235294f, 0.4f, 0.4862745f, 0.49411765f, 0.40784314f, 0.4862745f, 0.49411765f, 0.40784314f, 0.46666667f, 0.48235294f, 0.39215687f, 0.45490196f, 0.47058824f, 0.38039216f, 0.40784314f, 0.42745098f, 0.3372549f, 0.43529412f, 0.45490196f, 0.3647059f, 0.49803922f, 0.5294118f, 0.43529412f, 0.5294118f, 0.56078434f, 0.46666667f, 0.52156866f, 0.56078434f, 0.4627451f, 0.5372549f, 0.5764706f, 0.47843137f, 0.50980395f, 0.56078434f, 0.45882353f, 0.4392157f, 0.47843137f, 0.38039216f, 0.4f, 0.43529412f, 0.32941177f, 0.47058824f, 0.5058824f, 0.4f, 0.50980395f, 0.54509807f, 0.4392157f, 0.47843137f, 0.5137255f, 0.40784314f, 0.44313726f, 0.47843137f, 0.37254903f, 0.44313726f, 0.47843137f, 0.37254903f, 0.44313726f, 0.47843137f, 0.37254903f, 0.42745098f, 0.4627451f, 0.35686275f, 0.43137255f, 0.46666667f, 0.36078432f, 0.4862745f, 0.52156866f, 0.41568628f, 0.45490196f, 0.49019608f, 0.38431373f, 0.31764707f, 0.3529412f, 0.24705882f, 0.34509805f, 0.38039216f, 0.27450982f, 0.41960785f, 0.45490196f, 0.34901962f, 0.39607844f, 0.43137255f, 0.3254902f, 0.47843137f, 0.5137255f, 0.40784314f, 0.38431373f, 0.3764706f, 0.29411766f, 0.44705883f, 0.4392157f, 0.35686275f, 0.50980395f, 0.5058824f, 0.42352942f, 0.5294118f, 0.5254902f, 0.44313726f, 0.48235294f, 0.49019608f, 0.40392157f, 0.43137255f, 0.4392157f, 0.3529412f, 0.4117647f, 0.42745098f, 0.3372549f, 0.42745098f, 0.44313726f, 0.3529412f, 0.39215687f, 0.4117647f, 0.32156864f, 0.41568628f, 0.43529412f, 0.34509805f, 0.46666667f, 0.49803922f, 0.40392157f, 0.49019608f, 0.52156866f, 0.42745098f, 0.4627451f, 0.5019608f, 0.40392157f, 0.48235294f, 0.52156866f, 0.42352942f, 0.49411765f, 0.54509807f, 0.44313726f, 0.45490196f, 0.5058824f, 0.40392157f, 0.50980395f, 0.54509807f, 0.4392157f, 0.49803922f, 0.53333336f, 0.42745098f, 0.4745098f, 0.50980395f, 0.40392157f, 0.45882353f, 0.49411765f, 0.3882353f, 0.4745098f, 0.50980395f, 0.40392157f, 0.50980395f, 0.54509807f, 0.4392157f, 0.53333336f, 0.5686275f, 0.4627451f, 0.5294118f, 0.5647059f, 0.45882353f, 0.45882353f, 0.49411765f, 0.3882353f, 0.4745098f, 0.50980395f, 0.40392157f, 0.47843137f, 0.5137255f, 0.40784314f, 0.38431373f, 0.41960785f, 0.3137255f, 0.39607844f, 0.43137255f, 0.3254902f, 0.45490196f, 0.49019608f, 0.38431373f, 0.4f, 0.43529412f, 0.32941177f, 0.40784314f, 0.44313726f, 0.3372549f, 0.4f, 0.39215687f, 0.30980393f, 0.41568628f, 0.40784314f, 0.3254902f, 0.42352942f, 0.41960785f, 0.3372549f, 0.41568628f, 0.4117647f, 0.32941177f, 0.36862746f, 0.3764706f, 0.2901961f, 0.34901962f, 0.35686275f, 0.27058825f, 0.39215687f, 0.40784314f, 0.31764707f, 0.45490196f, 0.47058824f, 0.38039216f, 0.41568628f, 0.43529412f, 0.34509805f, 0.43137255f, 0.4509804f, 0.36078432f, 0.47843137f, 0.50980395f, 0.41568628f, 0.49411765f, 0.5254902f, 0.43137255f, 0.45490196f, 0.49411765f, 0.39607844f, 0.49019608f, 0.5294118f, 0.43137255f, 0.54901963f, 0.6f, 0.49803922f, 0.5647059f, 0.6156863f, 0.5137255f, 0.54509807f, 0.5803922f, 0.4745098f, 0.44313726f, 0.47843137f, 0.37254903f, 0.36078432f, 0.39607844f, 0.2901961f, 0.39607844f, 0.43137255f, 0.3254902f, 0.5019608f, 0.5372549f, 0.43137255f, 0.5803922f, 0.6156863f, 0.50980395f, 0.5882353f, 0.62352943f, 0.5176471f, 0.5686275f, 0.6039216f, 0.49803922f, 0.41568628f, 0.4509804f, 0.34509805f, 0.29411766f, 0.32941177f, 0.22352941f, 0.34117648f, 0.3764706f, 0.27058825f, 0.33333334f, 0.36862746f, 0.2627451f, 0.32941177f, 0.3647059f, 0.25882354f, 0.4627451f, 0.49803922f, 0.39215687f, 0.46666667f, 0.5019608f, 0.39607844f, 0.36078432f, 0.39607844f, 0.2901961f, 0.5647059f, 0.5568628f, 0.4745098f, 0.44705883f, 0.4392157f, 0.35686275f, 0.30980393f, 0.30588236f, 0.22352941f, 0.23137255f, 0.22745098f, 0.14509805f, 0.21176471f, 0.21960784f, 0.13333334f, 0.25882354f, 0.26666668f, 0.18039216f, 0.36078432f, 0.3764706f, 0.28627452f, 0.45490196f, 0.47058824f, 0.38039216f, 0.45490196f, 0.4745098f, 0.38431373f, 0.4509804f, 0.47058824f, 0.38039216f, 0.47058824f, 0.5019608f, 0.40784314f, 0.4627451f, 0.49411765f, 0.4f, 0.40392157f, 0.44313726f, 0.34509805f, 0.41568628f, 0.45490196f, 0.35686275f, 0.46666667f, 0.5176471f, 0.41568628f, 0.4862745f, 0.5372549f, 0.43529412f, 0.41568628f, 0.45490196f, 0.34901962f, 0.3764706f, 0.4117647f, 0.30588236f, 0.3764706f, 0.4117647f, 0.30588236f, 0.4392157f, 0.4745098f, 0.36862746f, 0.49411765f, 0.5294118f, 0.42352942f, 0.5176471f, 0.5529412f, 0.44705883f, 0.54509807f, 0.5803922f, 0.4745098f, 0.5882353f, 0.62352943f, 0.5176471f, 0.4862745f, 0.52156866f, 0.41568628f, 0.28235295f, 0.31764707f, 0.21176471f, 0.30588236f, 0.34117648f, 0.23529412f, 0.3019608f, 0.3372549f, 0.23137255f, 0.28235295f, 0.31764707f, 0.21176471f, 0.4627451f, 0.49803922f, 0.39215687f, 0.5294118f, 0.5647059f, 0.45882353f, 0.42352942f, 0.45882353f, 0.3529412f, 0.63529414f, 0.627451f, 0.54509807f, 0.50980395f, 0.5019608f, 0.41960785f, 0.36078432f, 0.35686275f, 0.27450982f, 0.27450982f, 0.27058825f, 0.1882353f, 0.23529412f, 0.24313726f, 0.15686275f, 0.27450982f, 0.28235295f, 0.19607843f, 0.38431373f, 0.4f, 0.30980393f, 0.49803922f, 0.5137255f, 0.42352942f, 0.5137255f, 0.53333336f, 0.44313726f, 0.4627451f, 0.48235294f, 0.39215687f, 0.4392157f, 0.47058824f, 0.3764706f, 0.40784314f, 0.4392157f, 0.34509805f, 0.31764707f, 0.35686275f, 0.25882354f, 0.26666668f, 0.30588236f, 0.20784314f, 0.23921569f, 0.2901961f, 0.1882353f, 0.20392157f, 0.25490198f, 0.15294118f, 0.3137255f, 0.3529412f, 0.24705882f, 0.3254902f, 0.3647059f, 0.25882354f, 0.4f, 0.4392157f, 0.33333334f, 0.49019608f, 0.5294118f, 0.42352942f, 0.49803922f, 0.5372549f, 0.43137255f, 0.45882353f, 0.49803922f, 0.39215687f, 0.48235294f, 0.52156866f, 0.41568628f, 0.5529412f, 0.5921569f, 0.4862745f, 0.5529412f, 0.5921569f, 0.4862745f, 0.4745098f, 0.5137255f, 0.40784314f, 0.5254902f, 0.5647059f, 0.45882353f, 0.49803922f, 0.5372549f, 0.43137255f, 0.47058824f, 0.50980395f, 0.40392157f, 0.53333336f, 0.57254905f, 0.46666667f, 0.5019608f, 0.5411765f, 0.43529412f, 0.44705883f, 0.4862745f, 0.38039216f, 0.54901963f, 0.5411765f, 0.45882353f, 0.53333336f, 0.5254902f, 0.44313726f, 0.50980395f, 0.5058824f, 0.42352942f, 0.47058824f, 0.46666667f, 0.38431373f, 0.4f, 0.40784314f, 0.32156864f, 0.3882353f, 0.39607844f, 0.30980393f, 0.4862745f, 0.5019608f, 0.4117647f, 0.6156863f, 0.6313726f, 0.5411765f, 0.5019608f, 0.52156866f, 0.43137255f, 0.44313726f, 0.4627451f, 0.37254903f, 0.43137255f, 0.4627451f, 0.36862746f, 0.44313726f, 0.4745098f, 0.38039216f, 0.39607844f, 0.43529412f, 0.3372549f, 0.34901962f, 0.3882353f, 0.2901961f, 0.2901961f, 0.34117648f, 0.23921569f, 0.21960784f, 0.27058825f, 0.16862746f, 0.32941177f, 0.36862746f, 0.2627451f, 0.28235295f, 0.32156864f, 0.21568628f, 0.32156864f, 0.36078432f, 0.25490198f, 0.45882353f, 0.49803922f, 0.39215687f, 0.5411765f, 0.5803922f, 0.4745098f, 0.5058824f, 0.54509807f, 0.4392157f, 0.45882353f, 0.49803922f, 0.39215687f, 0.45882353f, 0.49803922f, 0.39215687f, 0.5137255f, 0.5529412f, 0.44705883f, 0.5568628f, 0.59607846f, 0.49019608f, 0.57254905f, 0.6117647f, 0.5058824f, 0.47843137f, 0.5176471f, 0.4117647f, 0.4627451f, 0.5019608f, 0.39607844f, 0.44705883f, 0.4862745f, 0.38039216f, 0.37254903f, 0.4117647f, 0.30588236f, 0.4392157f, 0.47843137f, 0.37254903f, 0.4509804f, 0.44313726f, 0.36078432f, 0.5294118f, 0.52156866f, 0.4392157f, 0.45882353f, 0.45490196f, 0.37254903f, 0.4509804f, 0.44705883f, 0.3647059f, 0.48235294f, 0.49019608f, 0.40392157f, 0.49803922f, 0.5058824f, 0.41960785f, 0.5529412f, 0.5686275f, 0.47843137f, 0.5647059f, 0.5803922f, 0.49019608f, 0.24705882f, 0.26666668f, 0.1764706f, 0.27450982f, 0.29411766f, 0.20392157f, 0.43529412f, 0.46666667f, 0.37254903f, 0.50980395f, 0.5411765f, 0.44705883f, 0.5176471f, 0.5568628f, 0.45882353f, 0.5372549f, 0.5764706f, 0.47843137f, 0.49803922f, 0.54901963f, 0.44705883f, 0.5254902f, 0.5764706f, 0.4745098f, 0.49803922f, 0.54901963f, 0.4392157f, 0.3372549f, 0.3882353f, 0.2784314f, 0.31764707f, 0.36862746f, 0.25882354f, 0.4117647f, 0.4627451f, 0.3529412f, 0.43137255f, 0.48235294f, 0.37254903f, 0.4392157f, 0.49019608f, 0.38039216f, 0.4392157f, 0.49019608f, 0.38039216f, 0.3882353f, 0.4392157f, 0.32941177f, 0.47843137f, 0.5294118f, 0.41960785f, 0.49411765f, 0.54509807f, 0.43529412f, 0.5137255f, 0.5647059f, 0.45490196f, 0.46666667f, 0.5176471f, 0.40784314f, 0.38431373f, 0.43529412f, 0.3254902f, 0.3882353f, 0.4392157f, 0.32941177f, 0.42352942f, 0.4745098f, 0.3647059f, 0.40784314f, 0.45882353f, 0.34901962f, 0.44313726f, 0.43529412f, 0.3529412f, 0.44313726f, 0.43529412f, 0.3529412f, 0.40784314f, 0.40392157f, 0.32156864f, 0.39607844f, 0.39215687f, 0.30980393f, 0.34901962f, 0.35686275f, 0.27058825f, 0.42745098f, 0.43529412f, 0.34901962f, 0.57254905f, 0.5882353f, 0.49803922f, 0.5254902f, 0.5411765f, 0.4509804f, 0.2901961f, 0.30980393f, 0.21960784f, 0.30588236f, 0.3254902f, 0.23529412f, 0.42352942f, 0.45490196f, 0.36078432f, 0.43137255f, 0.4627451f, 0.36862746f, 0.4f, 0.4392157f, 0.34117648f, 0.4392157f, 0.47843137f, 0.38039216f, 0.4627451f, 0.5137255f, 0.4117647f, 0.5372549f, 0.5882353f, 0.4862745f, 0.5372549f, 0.5882353f, 0.47843137f, 0.52156866f, 0.57254905f, 0.4627451f, 0.50980395f, 0.56078434f, 0.4509804f, 0.44705883f, 0.49803922f, 0.3882353f, 0.34901962f, 0.4f, 0.2901961f, 0.3254902f, 0.3764706f, 0.26666668f, 0.36078432f, 0.4117647f, 0.3019608f, 0.37254903f, 0.42352942f, 0.3137255f, 0.46666667f, 0.5176471f, 0.40784314f, 0.44313726f, 0.49411765f, 0.38431373f, 0.46666667f, 0.5176471f, 0.40784314f, 0.47843137f, 0.5294118f, 0.41960785f, 0.44705883f, 0.49803922f, 0.3882353f, 0.43529412f, 0.4862745f, 0.3764706f, 0.4117647f, 0.4627451f, 0.3529412f, 0.34901962f, 0.4f, 0.2901961f, 0.45882353f, 0.4509804f, 0.36862746f, 0.4f, 0.39215687f, 0.30980393f, 0.3764706f, 0.37254903f, 0.2901961f, 0.35686275f, 0.3529412f, 0.27058825f, 0.25882354f, 0.26666668f, 0.18039216f, 0.4f, 0.40784314f, 0.32156864f, 0.63529414f, 0.6509804f, 0.56078434f, 0.56078434f, 0.5764706f, 0.4862745f, 0.4627451f, 0.48235294f, 0.39215687f, 0.43529412f, 0.45490196f, 0.3647059f, 0.49411765f, 0.5254902f, 0.43137255f, 0.46666667f, 0.49803922f, 0.40392157f, 0.42352942f, 0.4627451f, 0.3647059f, 0.41568628f, 0.45490196f, 0.35686275f, 0.32941177f, 0.38039216f, 0.2784314f, 0.3019608f, 0.3529412f, 0.2509804f, 0.38039216f, 0.4392157f, 0.3254902f, 0.5137255f, 0.57254905f, 0.45882353f, 0.5411765f, 0.6f, 0.4862745f, 0.43529412f, 0.49411765f, 0.38039216f, 0.3372549f, 0.39607844f, 0.28235295f, 0.29411766f, 0.3529412f, 0.23921569f, 0.2784314f, 0.3372549f, 0.22352941f, 0.2901961f, 0.34901962f, 0.23529412f, 0.34117648f, 0.4f, 0.28627452f, 0.31764707f, 0.3764706f, 0.2627451f, 0.3647059f, 0.42352942f, 0.30980393f, 0.42352942f, 0.48235294f, 0.36862746f, 0.41568628f, 0.4745098f, 0.36078432f, 0.40784314f, 0.46666667f, 0.3529412f, 0.40784314f, 0.46666667f, 0.3529412f, 0.3764706f, 0.43529412f, 0.32156864f, 0.43137255f, 0.42352942f, 0.34117648f, 0.44705883f, 0.4392157f, 0.35686275f, 0.43137255f, 0.42745098f, 0.34509805f, 0.42745098f, 0.42352942f, 0.34117648f, 0.38039216f, 0.3882353f, 0.3019608f, 0.4745098f, 0.48235294f, 0.39607844f, 0.63529414f, 0.6509804f, 0.56078434f, 0.58431375f, 0.6f, 0.50980395f, 0.5058824f, 0.5254902f, 0.43529412f, 0.45882353f, 0.47843137f, 0.3882353f, 0.49019608f, 0.52156866f, 0.42745098f, 0.44705883f, 0.47843137f, 0.38431373f, 0.39215687f, 0.43137255f, 0.33333334f, 0.36078432f, 0.4f, 0.3019608f, 0.22745098f, 0.2784314f, 0.1764706f, 0.15294118f, 0.20392157f, 0.101960786f, 0.36078432f, 0.41960785f, 0.30588236f, 0.4509804f, 0.50980395f, 0.39607844f, 0.43529412f, 0.49411765f, 0.38039216f, 0.39215687f, 0.4509804f, 0.3372549f, 0.43137255f, 0.49019608f, 0.3764706f, 0.41568628f, 0.4745098f, 0.36078432f, 0.34509805f, 0.40392157f, 0.2901961f, 0.33333334f, 0.39215687f, 0.2784314f, 0.31764707f, 0.3764706f, 0.2627451f, 0.3137255f, 0.37254903f, 0.25882354f, 0.38039216f, 0.4392157f, 0.3254902f, 0.41568628f, 0.4745098f, 0.36078432f, 0.3647059f, 0.42352942f, 0.30980393f, 0.35686275f, 0.41568628f, 0.3019608f, 0.42352942f, 0.48235294f, 0.36862746f, 0.47843137f, 0.5372549f, 0.42352942f, 0.39215687f, 0.38431373f, 0.3019608f, 0.5058824f, 0.49803922f, 0.41568628f, 0.44313726f, 0.4392157f, 0.35686275f, 0.4392157f, 0.43529412f, 0.3529412f, 0.49019608f, 0.49803922f, 0.4117647f, 0.4745098f, 0.48235294f, 0.39607844f, 0.49803922f, 0.5137255f, 0.42352942f, 0.52156866f, 0.5372549f, 0.44705883f, 0.49019608f, 0.50980395f, 0.41960785f, 0.44313726f, 0.4627451f, 0.37254903f, 0.47058824f, 0.5019608f, 0.40784314f, 0.40784314f, 0.4392157f, 0.34509805f, 0.3529412f, 0.39215687f, 0.29411766f, 0.39215687f, 0.43137255f, 0.33333334f, 0.4f, 0.4509804f, 0.34901962f, 0.4509804f, 0.5019608f, 0.4f, 0.54901963f, 0.60784316f, 0.49411765f, 0.47843137f, 0.5411765f, 0.42745098f, 0.35686275f, 0.41960785f, 0.30588236f, 0.3529412f, 0.41568628f, 0.3019608f, 0.4627451f, 0.5254902f, 0.4117647f, 0.4745098f, 0.5372549f, 0.42352942f, 0.43529412f, 0.49803922f, 0.38431373f, 0.4745098f, 0.5372549f, 0.42352942f, 0.42352942f, 0.4862745f, 0.37254903f, 0.4117647f, 0.4745098f, 0.36078432f, 0.44705883f, 0.50980395f, 0.39607844f, 0.4509804f, 0.5137255f, 0.4f, 0.38431373f, 0.44705883f, 0.33333334f, 0.37254903f, 0.43529412f, 0.32156864f, 0.43137255f, 0.49411765f, 0.38039216f, 0.46666667f, 0.5294118f, 0.41568628f, 0.4392157f, 0.43137255f, 0.34901962f, 0.5019608f, 0.49411765f, 0.4117647f, 0.3254902f, 0.32156864f, 0.23921569f, 0.28235295f, 0.2784314f, 0.19607843f, 0.3764706f, 0.38431373f, 0.29803923f, 0.33333334f, 0.34117648f, 0.25490198f, 0.36078432f, 0.3764706f, 0.28627452f, 0.49411765f, 0.50980395f, 0.41960785f, 0.5411765f, 0.56078434f, 0.47058824f, 0.4509804f, 0.47058824f, 0.38039216f, 0.45490196f, 0.4862745f, 0.39215687f, 0.43137255f, 0.4627451f, 0.36862746f, 0.42352942f, 0.4627451f, 0.3647059f, 0.49411765f, 0.53333336f, 0.43529412f, 0.5294118f, 0.5803922f, 0.47843137f, 0.6f, 0.654902f, 0.5529412f, 0.48235294f, 0.54509807f, 0.43137255f, 0.3764706f, 0.4392157f, 0.3254902f, 0.2901961f, 0.3529412f, 0.23921569f, 0.3254902f, 0.3882353f, 0.27450982f, 0.3882353f, 0.4509804f, 0.3372549f, 0.37254903f, 0.43529412f, 0.32156864f, 0.38039216f, 0.44313726f, 0.32941177f, 0.45882353f, 0.52156866f, 0.40784314f, 0.40392157f, 0.46666667f, 0.3529412f, 0.3764706f, 0.4392157f, 0.3254902f, 0.39215687f, 0.45490196f, 0.34117648f, 0.4117647f, 0.4745098f, 0.36078432f, 0.40392157f, 0.46666667f, 0.3529412f, 0.41960785f, 0.48235294f, 0.36862746f, 0.4117647f, 0.4745098f, 0.36078432f, 0.34509805f, 0.40784314f, 0.29411766f, 0.4627451f, 0.45490196f, 0.37254903f, 0.41960785f, 0.4117647f, 0.32941177f, 0.26666668f, 0.2627451f, 0.18039216f, 0.29411766f, 0.2901961f, 0.20784314f, 0.3529412f, 0.36078432f, 0.27450982f, 0.30980393f, 0.31764707f, 0.23137255f, 0.3882353f, 0.40392157f, 0.3137255f, 0.5137255f, 0.5294118f, 0.4392157f, 0.5411765f, 0.56078434f, 0.47058824f, 0.4f, 0.41960785f, 0.32941177f, 0.39607844f, 0.42745098f, 0.33333334f, 0.42745098f, 0.45882353f, 0.3647059f, 0.4509804f, 0.49019608f, 0.39215687f, 0.4745098f, 0.5137255f, 0.41568628f, 0.41568628f, 0.46666667f, 0.3647059f, 0.42352942f, 0.47843137f, 0.3764706f, 0.3529412f, 0.41568628f, 0.3019608f, 0.27058825f, 0.34509805f, 0.22745098f, 0.2784314f, 0.3529412f, 0.23529412f, 0.3372549f, 0.4117647f, 0.29411766f, 0.32941177f, 0.40392157f, 0.28627452f, 0.3137255f, 0.3882353f, 0.27058825f, 0.35686275f, 0.43137255f, 0.3137255f, 0.40392157f, 0.47843137f, 0.36078432f, 0.32941177f, 0.40392157f, 0.28627452f, 0.3372549f, 0.4117647f, 0.29411766f, 0.3647059f, 0.4392157f, 0.32156864f, 0.3764706f, 0.4509804f, 0.33333334f, 0.39607844f, 0.47058824f, 0.3529412f, 0.4509804f, 0.5254902f, 0.40784314f, 0.42745098f, 0.5019608f, 0.38431373f, 0.31764707f, 0.39215687f, 0.27450982f, 0.3882353f, 0.3764706f, 0.3019608f, 0.31764707f, 0.3137255f, 0.23137255f, 0.34901962f, 0.34509805f, 0.26666668f, 0.5294118f, 0.5372549f, 0.4509804f, 0.5411765f, 0.5568628f, 0.46666667f, 0.45490196f, 0.47058824f, 0.38039216f, 0.49019608f, 0.50980395f, 0.41960785f, 0.48235294f, 0.5137255f, 0.41960785f, 0.5411765f, 0.57254905f, 0.47843137f, 0.40784314f, 0.44705883f, 0.34901962f, 0.42745098f, 0.46666667f, 0.36862746f, 0.4509804f, 0.5019608f, 0.4f, 0.4509804f, 0.5019608f, 0.4f, 0.43529412f, 0.4862745f, 0.38431373f, 0.3882353f, 0.44313726f, 0.34117648f, 0.43529412f, 0.49019608f, 0.3882353f, 0.46666667f, 0.5411765f, 0.42352942f, 0.37254903f, 0.44705883f, 0.32941177f, 0.37254903f, 0.44705883f, 0.32941177f, 0.39215687f, 0.46666667f, 0.34901962f, 0.34901962f, 0.42352942f, 0.30588236f, 0.39215687f, 0.46666667f, 0.34901962f, 0.48235294f, 0.5568628f, 0.4392157f, 0.49411765f, 0.5686275f, 0.4509804f, 0.3882353f, 0.4627451f, 0.34509805f, 0.44705883f, 0.52156866f, 0.40392157f, 0.49411765f, 0.5686275f, 0.4509804f, 0.45490196f, 0.5294118f, 0.4117647f, 0.42352942f, 0.49803922f, 0.38039216f, 0.48235294f, 0.5568628f, 0.4392157f, 0.49803922f, 0.57254905f, 0.45490196f, 0.40784314f, 0.48235294f, 0.3647059f, 0.43137255f, 0.42745098f, 0.35686275f, 0.34117648f, 0.34901962f, 0.26666668f, 0.36078432f, 0.37254903f, 0.29803923f, 0.45490196f, 0.4745098f, 0.3882353f, 0.5019608f, 0.53333336f, 0.44313726f, 0.5411765f, 0.5803922f, 0.4862745f, 0.5411765f, 0.5882353f, 0.49411765f, 0.4862745f, 0.5411765f, 0.44313726f, 0.5568628f, 0.6117647f, 0.50980395f, 0.59607846f, 0.65882355f, 0.5529412f, 0.4745098f, 0.5372549f, 0.43137255f, 0.36862746f, 0.44313726f, 0.33333334f, 0.42352942f, 0.49803922f, 0.38039216f, 0.39215687f, 0.46666667f, 0.34901962f, 0.35686275f, 0.41960785f, 0.30588236f, 0.44705883f, 0.52156866f, 0.40392157f, 0.41960785f, 0.49411765f, 0.3764706f, 0.40392157f, 0.4862745f, 0.3647059f, 0.3882353f, 0.47058824f, 0.34901962f, 0.38431373f, 0.46666667f, 0.34509805f, 0.40784314f, 0.49019608f, 0.36862746f, 0.4392157f, 0.52156866f, 0.4f, 0.44313726f, 0.5254902f, 0.40392157f, 0.42745098f, 0.50980395f, 0.3882353f, 0.42745098f, 0.50980395f, 0.3882353f, 0.5137255f, 0.59607846f, 0.4745098f, 0.3764706f, 0.45882353f, 0.3372549f, 0.46666667f, 0.54901963f, 0.42745098f, 0.5058824f, 0.5882353f, 0.46666667f, 0.35686275f, 0.4392157f, 0.31764707f, 0.3882353f, 0.47058824f, 0.34901962f, 0.36862746f, 0.4509804f, 0.32941177f, 0.41568628f, 0.41960785f, 0.34901962f, 0.34509805f, 0.35686275f, 0.28235295f, 0.3137255f, 0.3254902f, 0.2509804f, 0.30980393f, 0.34117648f, 0.25882354f, 0.35686275f, 0.3882353f, 0.29803923f, 0.45882353f, 0.5058824f, 0.4117647f, 0.52156866f, 0.5764706f, 0.47843137f, 0.47843137f, 0.5411765f, 0.4392157f, 0.45490196f, 0.5294118f, 0.41960785f, 0.43529412f, 0.50980395f, 0.4f, 0.3529412f, 0.43137255f, 0.32156864f, 0.3764706f, 0.45490196f, 0.34509805f, 0.49019608f, 0.57254905f, 0.4509804f, 0.45490196f, 0.5372549f, 0.41568628f, 0.32156864f, 0.39607844f, 0.2784314f, 0.2784314f, 0.3529412f, 0.23529412f, 0.34117648f, 0.42352942f, 0.3019608f, 0.3764706f, 0.45882353f, 0.3372549f, 0.4f, 0.48235294f, 0.36078432f, 0.39607844f, 0.47843137f, 0.35686275f, 0.38039216f, 0.4627451f, 0.34117648f, 0.36862746f, 0.4509804f, 0.32941177f, 0.3529412f, 0.43529412f, 0.3137255f, 0.34117648f, 0.42352942f, 0.3019608f, 0.43529412f, 0.5176471f, 0.39607844f, 0.5019608f, 0.58431375f, 0.4627451f, 0.36078432f, 0.44313726f, 0.32156864f, 0.39607844f, 0.47843137f, 0.35686275f, 0.42745098f, 0.50980395f, 0.3882353f, 0.38039216f, 0.4627451f, 0.34117648f, 0.47058824f, 0.5529412f, 0.43137255f, 0.41960785f, 0.5019608f, 0.38039216f, 0.44313726f, 0.44705883f, 0.3764706f, 0.4509804f, 0.4627451f, 0.3882353f, 0.4117647f, 0.42352942f, 0.34901962f, 0.28627452f, 0.31764707f, 0.23529412f, 0.22352941f, 0.25490198f, 0.16470589f, 0.32156864f, 0.36862746f, 0.27450982f, 0.45882353f, 0.5137255f, 0.41568628f, 0.49803922f, 0.56078434f, 0.45882353f, 0.4117647f, 0.4862745f, 0.3764706f, 0.2784314f, 0.3529412f, 0.24313726f, 0.27058825f, 0.34901962f, 0.23921569f, 0.44705883f, 0.5254902f, 0.41568628f, 0.5686275f, 0.6509804f, 0.5294118f, 0.49411765f, 0.5764706f, 0.45490196f, 0.35686275f, 0.43137255f, 0.3137255f, 0.29411766f, 0.36862746f, 0.2509804f, 0.4f, 0.48235294f, 0.36078432f, 0.40784314f, 0.49019608f, 0.36862746f, 0.4117647f, 0.49411765f, 0.37254903f, 0.40392157f, 0.4862745f, 0.3647059f, 0.40784314f, 0.49019608f, 0.36862746f, 0.41568628f, 0.49803922f, 0.3764706f, 0.40784314f, 0.49019608f, 0.36862746f, 0.39215687f, 0.4745098f, 0.3529412f, 0.47058824f, 0.5529412f, 0.43137255f, 0.53333336f, 0.6156863f, 0.49411765f, 0.43529412f, 0.5176471f, 0.39607844f, 0.40392157f, 0.4862745f, 0.3647059f, 0.34901962f, 0.43137255f, 0.30980393f, 0.37254903f, 0.45490196f, 0.33333334f, 0.5411765f, 0.62352943f, 0.5019608f, 0.49803922f, 0.5803922f, 0.45882353f, 0.41568628f, 0.41960785f, 0.34901962f, 0.4627451f, 0.4745098f, 0.4f, 0.4627451f, 0.4745098f, 0.4f, 0.35686275f, 0.3882353f, 0.30588236f, 0.27450982f, 0.30588236f, 0.21568628f, 0.30980393f, 0.35686275f, 0.2627451f, 0.43529412f, 0.49019608f, 0.39215687f, 0.5176471f, 0.5803922f, 0.47843137f, 0.49411765f, 0.5686275f, 0.45882353f, 0.28235295f, 0.35686275f, 0.24705882f, 0.29803923f, 0.3764706f, 0.26666668f, 0.5058824f, 0.58431375f, 0.4745098f, 0.5294118f, 0.6117647f, 0.49019608f, 0.41568628f, 0.49803922f, 0.3764706f, 0.3882353f, 0.4627451f, 0.34509805f, 0.41960785f, 0.49411765f, 0.3764706f, 0.5019608f, 0.58431375f, 0.4627451f, 0.46666667f, 0.54901963f, 0.42745098f, 0.42352942f, 0.5058824f, 0.38431373f, 0.40392157f, 0.4862745f, 0.3647059f, 0.4117647f, 0.49411765f, 0.37254903f, 0.4f, 0.48235294f, 0.36078432f, 0.3372549f, 0.41960785f, 0.29803923f, 0.27058825f, 0.3529412f, 0.23137255f, 0.4f, 0.48235294f, 0.36078432f, 0.4509804f, 0.53333336f, 0.4117647f, 0.4392157f, 0.52156866f, 0.4f, 0.40392157f, 0.4862745f, 0.3647059f, 0.2627451f, 0.34509805f, 0.22352941f, 0.2627451f, 0.34509805f, 0.22352941f, 0.4392157f, 0.52156866f, 0.4f, 0.43137255f, 0.5137255f, 0.39215687f, 0.41960785f, 0.42352942f, 0.3529412f, 0.4117647f, 0.42352942f, 0.34901962f, 0.43137255f, 0.44313726f, 0.36862746f, 0.45882353f, 0.49019608f, 0.40784314f, 0.48235294f, 0.5137255f, 0.42352942f, 0.4627451f, 0.50980395f, 0.41568628f, 0.47843137f, 0.53333336f, 0.43529412f, 0.5254902f, 0.5882353f, 0.4862745f, 0.5803922f, 0.654902f, 0.54509807f, 0.3882353f, 0.4627451f, 0.3529412f, 0.36078432f, 0.4392157f, 0.32941177f, 0.44313726f, 0.52156866f, 0.4117647f, 0.39607844f, 0.47843137f, 0.35686275f, 0.34509805f, 0.42745098f, 0.30588236f, 0.38431373f, 0.45882353f, 0.34117648f, 0.40392157f, 0.47843137f, 0.36078432f, 0.4627451f, 0.54509807f, 0.42352942f, 0.47058824f, 0.5647059f, 0.4392157f, 0.47843137f, 0.57254905f, 0.44705883f, 0.47058824f, 0.5647059f, 0.4392157f, 0.44705883f, 0.5411765f, 0.41568628f, 0.40392157f, 0.49803922f, 0.37254903f, 0.3372549f, 0.43137255f, 0.30588236f, 0.28235295f, 0.3764706f, 0.2509804f, 0.34509805f, 0.4392157f, 0.3137255f, 0.33333334f, 0.42745098f, 0.3019608f, 0.3764706f, 0.47058824f, 0.34509805f, 0.43137255f, 0.5254902f, 0.4f, 0.34509805f, 0.4392157f, 0.3137255f, 0.3019608f, 0.39607844f, 0.27058825f, 0.38039216f, 0.4745098f, 0.34901962f, 0.36862746f, 0.4627451f, 0.3372549f, 0.44313726f, 0.44705883f, 0.3764706f, 0.43137255f, 0.44313726f, 0.36862746f, 0.44313726f, 0.45490196f, 0.38039216f, 0.4745098f, 0.5058824f, 0.42352942f, 0.48235294f, 0.5137255f, 0.42352942f, 0.39215687f, 0.4392157f, 0.34509805f, 0.36078432f, 0.41568628f, 0.31764707f, 0.41960785f, 0.48235294f, 0.38039216f, 0.5254902f, 0.6f, 0.49019608f, 0.43137255f, 0.5058824f, 0.39607844f, 0.3764706f, 0.45490196f, 0.34509805f, 0.34901962f, 0.42745098f, 0.31764707f, 0.3372549f, 0.41960785f, 0.29803923f, 0.40784314f, 0.49019608f, 0.36862746f, 0.43137255f, 0.5058824f, 0.3882353f, 0.3372549f, 0.41960785f, 0.29803923f, 0.41568628f, 0.49803922f, 0.3764706f, 0.47058824f, 0.5647059f, 0.4392157f, 0.5058824f, 0.6f, 0.4745098f, 0.45882353f, 0.5529412f, 0.42745098f, 0.39215687f, 0.4862745f, 0.36078432f, 0.3882353f, 0.48235294f, 0.35686275f, 0.45490196f, 0.54901963f, 0.42352942f, 0.52156866f, 0.6156863f, 0.49019608f, 0.33333334f, 0.42745098f, 0.3019608f, 0.27450982f, 0.36862746f, 0.24313726f, 0.31764707f, 0.4117647f, 0.28627452f, 0.4392157f, 0.53333336f, 0.40784314f, 0.4862745f, 0.5803922f, 0.45490196f, 0.4509804f, 0.54509807f, 0.41960785f, 0.4f, 0.49411765f, 0.36862746f, 0.34901962f, 0.44313726f, 0.31764707f, 0.44313726f, 0.44705883f, 0.3764706f, 0.46666667f, 0.47843137f, 0.40392157f, 0.43137255f, 0.44313726f, 0.36862746f, 0.34901962f, 0.38039216f, 0.29803923f, 0.2901961f, 0.32156864f, 0.23137255f, 0.21960784f, 0.26666668f, 0.17254902f, 0.24313726f, 0.29803923f, 0.2f, 0.3529412f, 0.41568628f, 0.3137255f, 0.4509804f, 0.5254902f, 0.41568628f, 0.43529412f, 0.50980395f, 0.4f, 0.4f, 0.47843137f, 0.36862746f, 0.3647059f, 0.44313726f, 0.33333334f, 0.39215687f, 0.4745098f, 0.3529412f, 0.45882353f, 0.5411765f, 0.41960785f, 0.45882353f, 0.53333336f, 0.41568628f, 0.38039216f, 0.4627451f, 0.34117648f, 0.4392157f, 0.52156866f, 0.4f, 0.46666667f, 0.56078434f, 0.43529412f, 0.45490196f, 0.54901963f, 0.42352942f, 0.37254903f, 0.46666667f, 0.34117648f, 0.29411766f, 0.3882353f, 0.2627451f, 0.3019608f, 0.39607844f, 0.27058825f, 0.4f, 0.49411765f, 0.36862746f, 0.4862745f, 0.5803922f, 0.45490196f, 0.29803923f, 0.39215687f, 0.26666668f, 0.29411766f, 0.3882353f, 0.2627451f, 0.32156864f, 0.41568628f, 0.2901961f, 0.3764706f, 0.47058824f, 0.34509805f, 0.45490196f, 0.54901963f, 0.42352942f, 0.44313726f, 0.5372549f, 0.4117647f, 0.3372549f, 0.43137255f, 0.30588236f, 0.3019608f, 0.39607844f, 0.27058825f, 0.5019608f, 0.5058824f, 0.43529412f, 0.5137255f, 0.5254902f, 0.4509804f, 0.4117647f, 0.42352942f, 0.34901962f, 0.27058825f, 0.3019608f, 0.21960784f, 0.25882354f, 0.2901961f, 0.2f, 0.29803923f, 0.34509805f, 0.2509804f, 0.38431373f, 0.4392157f, 0.34117648f, 0.49803922f, 0.56078434f, 0.45882353f, 0.4627451f, 0.5372549f, 0.42745098f, 0.45882353f, 0.53333336f, 0.42352942f, 0.44705883f, 0.5254902f, 0.41568628f, 0.4509804f, 0.5294118f, 0.41960785f, 0.44313726f, 0.5254902f, 0.40392157f, 0.4f, 0.48235294f, 0.36078432f, 0.4f, 0.4745098f, 0.35686275f, 0.44313726f, 0.5254902f, 0.40392157f, 0.41568628f, 0.50980395f, 0.38431373f, 0.4392157f, 0.53333336f, 0.40784314f, 0.4509804f, 0.54509807f, 0.41960785f, 0.43529412f, 0.5294118f, 0.40392157f, 0.41960785f, 0.5137255f, 0.3882353f, 0.41568628f, 0.50980395f, 0.38431373f, 0.40784314f, 0.5019608f, 0.3764706f, 0.39607844f, 0.49019608f, 0.3647059f, 0.35686275f, 0.4509804f, 0.3254902f, 0.45882353f, 0.5529412f, 0.42745098f, 0.49019608f, 0.58431375f, 0.45882353f, 0.41960785f, 0.5137255f, 0.3882353f, 0.43529412f, 0.5294118f, 0.40392157f, 0.41960785f, 0.5137255f, 0.3882353f, 0.32941177f, 0.42352942f, 0.29803923f, 0.34901962f, 0.44313726f, 0.31764707f, 0.40392157f, 0.40784314f, 0.3372549f, 0.3882353f, 0.4f, 0.3254902f, 0.36862746f, 0.38039216f, 0.30588236f, 0.37254903f, 0.40392157f, 0.32156864f, 0.45882353f, 0.49019608f, 0.4f, 0.5294118f, 0.5764706f, 0.48235294f, 0.49803922f, 0.5529412f, 0.45490196f, 0.40784314f, 0.47058824f, 0.36862746f, 0.3372549f, 0.4117647f, 0.3019608f, 0.25882354f, 0.33333334f, 0.22352941f, 0.30588236f, 0.38431373f, 0.27450982f, 0.4862745f, 0.5647059f, 0.45490196f, 0.5647059f, 0.64705884f, 0.5254902f, 0.4745098f, 0.5568628f, 0.43529412f, 0.4117647f, 0.4862745f, 0.36862746f, 0.43137255f, 0.5137255f, 0.39215687f, 0.39607844f, 0.49019608f, 0.3647059f, 0.41568628f, 0.5137255f, 0.3882353f, 0.41960785f, 0.5176471f, 0.39215687f, 0.47843137f, 0.5764706f, 0.4509804f, 0.5372549f, 0.63529414f, 0.50980395f, 0.44313726f, 0.5411765f, 0.41568628f, 0.34901962f, 0.44705883f, 0.32156864f, 0.39215687f, 0.49019608f, 0.3647059f, 0.3647059f, 0.4627451f, 0.3372549f, 0.4627451f, 0.56078434f, 0.43529412f, 0.47058824f, 0.5686275f, 0.44313726f, 0.3764706f, 0.4745098f, 0.34901962f, 0.3529412f, 0.4509804f, 0.3254902f, 0.42352942f, 0.52156866f, 0.39607844f, 0.43529412f, 0.53333336f, 0.40784314f, 0.36862746f, 0.46666667f, 0.34117648f, 0.39215687f, 0.39607844f, 0.3254902f, 0.3137255f, 0.3254902f, 0.2509804f, 0.3019608f, 0.3137255f, 0.23921569f, 0.39215687f, 0.42352942f, 0.34117648f, 0.50980395f, 0.5411765f, 0.4509804f, 0.5137255f, 0.56078434f, 0.46666667f, 0.4392157f, 0.49411765f, 0.39607844f, 0.3647059f, 0.42745098f, 0.3254902f, 0.2627451f, 0.3372549f, 0.22745098f, 0.20392157f, 0.2784314f, 0.16862746f, 0.22745098f, 0.30588236f, 0.19607843f, 0.3764706f, 0.45490196f, 0.34509805f, 0.49803922f, 0.5803922f, 0.45882353f, 0.5019608f, 0.58431375f, 0.4627451f, 0.45882353f, 0.53333336f, 0.41568628f, 0.43529412f, 0.5176471f, 0.39607844f, 0.43137255f, 0.5254902f, 0.4f, 0.43529412f, 0.53333336f, 0.40784314f, 0.42352942f, 0.52156866f, 0.39607844f, 0.43529412f, 0.53333336f, 0.40784314f, 0.47058824f, 0.5686275f, 0.44313726f, 0.4392157f, 0.5372549f, 0.4117647f, 0.37254903f, 0.47058824f, 0.34509805f, 0.34509805f, 0.44313726f, 0.31764707f, 0.34117648f, 0.4392157f, 0.3137255f, 0.4117647f, 0.50980395f, 0.38431373f, 0.4509804f, 0.54901963f, 0.42352942f, 0.4117647f, 0.50980395f, 0.38431373f, 0.36078432f, 0.45882353f, 0.33333334f, 0.35686275f, 0.45490196f, 0.32941177f, 0.3882353f, 0.4862745f, 0.36078432f, 0.40784314f, 0.5058824f, 0.38039216f, 0.5058824f, 0.50980395f, 0.4392157f, 0.38431373f, 0.39607844f, 0.32156864f, 0.36078432f, 0.37254903f, 0.29803923f, 0.46666667f, 0.49803922f, 0.41568628f, 0.5372549f, 0.5686275f, 0.47843137f, 0.47058824f, 0.5176471f, 0.42352942f, 0.40784314f, 0.4627451f, 0.3647059f, 0.4117647f, 0.4745098f, 0.37254903f, 0.43529412f, 0.50980395f, 0.4f, 0.43137255f, 0.5058824f, 0.39607844f, 0.43137255f, 0.50980395f, 0.4f, 0.4509804f, 0.5294118f, 0.41960785f, 0.47058824f, 0.5529412f, 0.43137255f, 0.46666667f, 0.54901963f, 0.42745098f, 0.44705883f, 0.52156866f, 0.40392157f, 0.42352942f, 0.5058824f, 0.38431373f, 0.4f, 0.49411765f, 0.36862746f, 0.42352942f, 0.52156866f, 0.39607844f, 0.4509804f, 0.54901963f, 0.42352942f, 0.41960785f, 0.5176471f, 0.39215687f, 0.38039216f, 0.47843137f, 0.3529412f, 0.40392157f, 0.5019608f, 0.3764706f, 0.41960785f, 0.5176471f, 0.39215687f, 0.38039216f, 0.47843137f, 0.3529412f, 0.37254903f, 0.47058824f, 0.34509805f, 0.38039216f, 0.47843137f, 0.3529412f, 0.39215687f, 0.49019608f, 0.3647059f, 0.4f, 0.49803922f, 0.37254903f, 0.4117647f, 0.50980395f, 0.38431373f, 0.41960785f, 0.5176471f, 0.39215687f, 0.41568628f, 0.5137255f, 0.3882353f, 0.40784314f, 0.5058824f, 0.38039216f, 0.57254905f, 0.5764706f, 0.5058824f, 0.4627451f, 0.4745098f, 0.4f, 0.40784314f, 0.41960785f, 0.34509805f, 0.4117647f, 0.44313726f, 0.36078432f, 0.3882353f, 0.41960785f, 0.32941177f, 0.3137255f, 0.36078432f, 0.26666668f, 0.33333334f, 0.3882353f, 0.2901961f, 0.41960785f, 0.48235294f, 0.38039216f, 0.4627451f, 0.5372549f, 0.42745098f, 0.46666667f, 0.5411765f, 0.43137255f, 0.47058824f, 0.54901963f, 0.4392157f, 0.48235294f, 0.56078434f, 0.4509804f, 0.4745098f, 0.5568628f, 0.43529412f, 0.45490196f, 0.5372549f, 0.41568628f, 0.45490196f, 0.5294118f, 0.4117647f, 0.45490196f, 0.5372549f, 0.41568628f, 0.35686275f, 0.4509804f, 0.3254902f, 0.39607844f, 0.49411765f, 0.36862746f, 0.4862745f, 0.58431375f, 0.45882353f, 0.45882353f, 0.5568628f, 0.43137255f, 0.3254902f, 0.42352942f, 0.29803923f, 0.32941177f, 0.42745098f, 0.3019608f, 0.41960785f, 0.5176471f, 0.39215687f, 0.4392157f, 0.5372549f, 0.4117647f, 0.43137255f, 0.5294118f, 0.40392157f, 0.42745098f, 0.5254902f, 0.4f, 0.39607844f, 0.49411765f, 0.36862746f, 0.37254903f, 0.47058824f, 0.34509805f, 0.4117647f, 0.50980395f, 0.38431373f, 0.46666667f, 0.5647059f, 0.4392157f, 0.43529412f, 0.53333336f, 0.40784314f, 0.35686275f, 0.45490196f, 0.32941177f, 0.5882353f, 0.5921569f, 0.52156866f, 0.5058824f, 0.5176471f, 0.44313726f, 0.4117647f, 0.42352942f, 0.34901962f, 0.34117648f, 0.37254903f, 0.2901961f, 0.31764707f, 0.34901962f, 0.25882354f, 0.32156864f, 0.36862746f, 0.27450982f, 0.37254903f, 0.42745098f, 0.32941177f, 0.42352942f, 0.4862745f, 0.38431373f, 0.40392157f, 0.47843137f, 0.36862746f, 0.34117648f, 0.41568628f, 0.30588236f, 0.3372549f, 0.41568628f, 0.30588236f, 0.43137255f, 0.50980395f, 0.4f, 0.5019608f, 0.58431375f, 0.4627451f, 0.49411765f, 0.5764706f, 0.45490196f, 0.47058824f, 0.54509807f, 0.42745098f, 0.4627451f, 0.54509807f, 0.42352942f, 0.36862746f, 0.46666667f, 0.34117648f, 0.34901962f, 0.45490196f, 0.3254902f, 0.43529412f, 0.5411765f, 0.4117647f, 0.4627451f, 0.5686275f, 0.4392157f, 0.34509805f, 0.4509804f, 0.32156864f, 0.30588236f, 0.4117647f, 0.28235295f, 0.3882353f, 0.49411765f, 0.3647059f, 0.4392157f, 0.54509807f, 0.41568628f, 0.42352942f, 0.5294118f, 0.4f, 0.41568628f, 0.52156866f, 0.39215687f, 0.38431373f, 0.49019608f, 0.36078432f, 0.3529412f, 0.45882353f, 0.32941177f, 0.34901962f, 0.45490196f, 0.3254902f, 0.36862746f, 0.4745098f, 0.34509805f, 0.36078432f, 0.46666667f, 0.3372549f, 0.33333334f, 0.4392157f, 0.30980393f, 0.5411765f, 0.54509807f, 0.4745098f, 0.48235294f, 0.49411765f, 0.41960785f, 0.39607844f, 0.40784314f, 0.33333334f, 0.34117648f, 0.37254903f, 0.2901961f, 0.39607844f, 0.42745098f, 0.3372549f, 0.46666667f, 0.5137255f, 0.41960785f, 0.47058824f, 0.5254902f, 0.42745098f, 0.42352942f, 0.4862745f, 0.38431373f, 0.43529412f, 0.50980395f, 0.4f, 0.3882353f, 0.4627451f, 0.3529412f, 0.39215687f, 0.47058824f, 0.36078432f, 0.4627451f, 0.5411765f, 0.43137255f, 0.4862745f, 0.5686275f, 0.44705883f, 0.44313726f, 0.5254902f, 0.40392157f, 0.42745098f, 0.5019608f, 0.38431373f, 0.4509804f, 0.53333336f, 0.4117647f, 0.3764706f, 0.4745098f, 0.34901962f, 0.30980393f, 0.41568628f, 0.28627452f, 0.33333334f, 0.4392157f, 0.30980393f, 0.40392157f, 0.50980395f, 0.38039216f, 0.4f, 0.5058824f, 0.3764706f, 0.38039216f, 0.4862745f, 0.35686275f, 0.41568628f, 0.52156866f, 0.39215687f, 0.45882353f, 0.5647059f, 0.43529412f, 0.39607844f, 0.5019608f, 0.37254903f, 0.2901961f, 0.39607844f, 0.26666668f, 0.25490198f, 0.36078432f, 0.23137255f, 0.32156864f, 0.42745098f, 0.29803923f, 0.36862746f, 0.4745098f, 0.34509805f, 0.3372549f, 0.44313726f, 0.3137255f, 0.32156864f, 0.42745098f, 0.29803923f, 0.3529412f, 0.45882353f, 0.32941177f, 0.43529412f, 0.4392157f, 0.36862746f, 0.41568628f, 0.42745098f, 0.3529412f, 0.38431373f, 0.39607844f, 0.32156864f, 0.36078432f, 0.39215687f, 0.30980393f, 0.4117647f, 0.44313726f, 0.3529412f, 0.45882353f, 0.5058824f, 0.4117647f, 0.43529412f, 0.49019608f, 0.39215687f, 0.36862746f, 0.43137255f, 0.32941177f, 0.41568628f, 0.49019608f, 0.38039216f, 0.45490196f, 0.5294118f, 0.41960785f, 0.48235294f, 0.56078434f, 0.4509804f, 0.45490196f, 0.53333336f, 0.42352942f, 0.39607844f, 0.47843137f, 0.35686275f, 0.3647059f, 0.44705883f, 0.3254902f, 0.40392157f, 0.47843137f, 0.36078432f, 0.45490196f, 0.5372549f, 0.41568628f, 0.41568628f, 0.5137255f, 0.3882353f, 0.4f, 0.5058824f, 0.3764706f, 0.36862746f, 0.4745098f, 0.34509805f, 0.3647059f, 0.47058824f, 0.34117648f, 0.39215687f, 0.49803922f, 0.36862746f, 0.39607844f, 0.5019608f, 0.37254903f, 0.41568628f, 0.52156866f, 0.39215687f, 0.48235294f, 0.5882353f, 0.45882353f, 0.44705883f, 0.5529412f, 0.42352942f, 0.3019608f, 0.40784314f, 0.2784314f, 0.2509804f, 0.35686275f, 0.22745098f, 0.35686275f, 0.4627451f, 0.33333334f, 0.44313726f, 0.54901963f, 0.41960785f, 0.4f, 0.5058824f, 0.3764706f, 0.34901962f, 0.45490196f, 0.3254902f, 0.3529412f, 0.45882353f, 0.32941177f, 0.44313726f, 0.44705883f, 0.3764706f, 0.47058824f, 0.48235294f, 0.40784314f, 0.49019608f, 0.5019608f, 0.42745098f, 0.47058824f, 0.5019608f, 0.41960785f, 0.44313726f, 0.4745098f, 0.38431373f, 0.41568628f, 0.4627451f, 0.36862746f, 0.4117647f, 0.46666667f, 0.36862746f, 0.40392157f, 0.46666667f, 0.3647059f, 0.47058824f, 0.54509807f, 0.43529412f, 0.5294118f, 0.6039216f, 0.49411765f, 0.52156866f, 0.6f, 0.49019608f, 0.42352942f, 0.5019608f, 0.39215687f, 0.35686275f, 0.4392157f, 0.31764707f, 0.3764706f, 0.45882353f, 0.3372549f, 0.41568628f, 0.49019608f, 0.37254903f, 0.40784314f, 0.49019608f, 0.36862746f, 0.5058824f, 0.6039216f, 0.47843137f, 0.58431375f, 0.6901961f, 0.56078434f, 0.5176471f, 0.62352943f, 0.49411765f, 0.38431373f, 0.49019608f, 0.36078432f, 0.33333334f, 0.4392157f, 0.30980393f, 0.31764707f, 0.42352942f, 0.29411766f, 0.34901962f, 0.45490196f, 0.3254902f, 0.4509804f, 0.5568628f, 0.42745098f, 0.5411765f, 0.64705884f, 0.5176471f, 0.47843137f, 0.58431375f, 0.45490196f, 0.44313726f, 0.54901963f, 0.41960785f, 0.4627451f, 0.5686275f, 0.4392157f, 0.46666667f, 0.57254905f, 0.44313726f, 0.4117647f, 0.5176471f, 0.3882353f, 0.34901962f, 0.45490196f, 0.3254902f, 0.32156864f, 0.42745098f, 0.29803923f,\n",
      "};\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def gen_X_test_hex(X_tst, num):\n",
    "    print('\\n X_test' + str(num) + '[]:\\n {')\n",
    "    c_str = 'const float x_test' + str(num) + '[32*32*3] = { '\n",
    "    for i in range(lenI):\n",
    "        for j in range(lenJ):\n",
    "            for k in range(lenK):\n",
    "                print(' ', X_tst[i][j][k], 'f,', sep='', end='')\n",
    "                c_str += \"%s%f%s\" % (' ', X_tst[i][j][k], 'f,')\n",
    "\n",
    "    c_str += '\\n};\\n'\n",
    "    print('\\n};\\n')\n",
    "    return c_str\n",
    "\n",
    "# Write X_test[0] input to a  header file\n",
    "index = 0\n",
    "file_name = \"cactus_X_test\" + str(index) + \"_data\" + '.h'\n",
    "print(file_name)\n",
    "with open(file_name, 'w') as file:\n",
    "    file.write(gen_X_test_hex(X_test[index], index))\n",
    "\n",
    "# Write X_test[2] input to a  header file\n",
    "index = 2\n",
    "file_name = \"cactus_X_test\" + str(index) + \"_data\" + '.h'\n",
    "print(file_name)\n",
    "with open(file_name, 'w') as file:\n",
    "    file.write(gen_X_test_hex(X_test[index], index))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "southern-greenhouse"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "name": "Ses15_cactus_seq_classify.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}