KerasModelv1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [],
   "source": [
    "#model v1 working\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Flatten, Conv2D, Activation, MaxPooling2D, Dropout\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "import cv2\n",
    "from tqdm import tqdm\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import os\n",
    "import cv2\n",
    "\n",
    "# from tensorflow import keras\n",
    "# from keras.tensorflow.keras import layers\n",
    "# from keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#LOADING OWN DATA\n",
    "\n",
    "DATADIR = r\"C:\\Users\\chris\\Downloads\\EaglesMain\"\n",
    "CATEGORIES = [\"Eagles\", \"NonEagles\"]\n",
    "\n",
    "for category in CATEGORIES:\n",
    "  path = os.path.join(DATADIR, category) #path to eagle or noneagle dir\n",
    "  for img in os.listdir(path):\n",
    "        img_array = cv2.imread(os.path.join(path,img), cv2.IMREAD_GRAYSCALE) \n",
    "        #note that I want to use colors in second version\n",
    "        plt.imshow(img_array, cmap=\"gray\" )\n",
    "        plt.show()\n",
    "        break\n",
    "  break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(169, 299)\n"
     ]
    }
   ],
   "source": [
    "print(img_array.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "IMG_SIZE = 70\n",
    "\n",
    "new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
    "plt.imshow(new_array, cmap = 'gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 91/91 [00:00<00:00, 1263.89it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████| 92/92 [00:00<00:00, 339.48it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "183\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "training_data = []\n",
    "\n",
    "def create_training_data():\n",
    "    for category in CATEGORIES:  # do dogs and cats\n",
    "\n",
    "        path = os.path.join(DATADIR,category)  # create path to Eagle and NonEagle\n",
    "        class_num = CATEGORIES.index(category)  # get the classification  (0 or a 1). 0=Eagle 1=NonEagle\n",
    "\n",
    "        for img in tqdm(os.listdir(path)):  # iterate over each image per dogs and cats\n",
    "            try:\n",
    "                img_array = cv2.imread(os.path.join(path,img) ,cv2.IMREAD_GRAYSCALE)  # convert to array\n",
    "                new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))  # resize to normalize data size\n",
    "                training_data.append([new_array, class_num])  # add this to our training_data\n",
    "            except Exception as e:  # in the interest in keeping the output clean...\n",
    "                pass\n",
    "            #except OSError as e:\n",
    "            #    print(\"OSErrroBad img most likely\", e, os.path.join(path,img))\n",
    "            #except Exception as e:\n",
    "            #    print(\"general exception\", e, os.path.join(path,img))\n",
    "\n",
    "create_training_data()\n",
    "\n",
    "print(len(training_data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "183\n",
      "169\n",
      "70\n"
     ]
    }
   ],
   "source": [
    "print(len(training_data))\n",
    "print(len(img_array))\n",
    "print(len(new_array))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "random.shuffle(training_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "1\n",
      "1\n",
      "0\n",
      "0\n",
      "1\n",
      "0\n",
      "0\n",
      "0\n"
     ]
    }
   ],
   "source": [
    "for sample in training_data[:10]:\n",
    "    print(sample[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = []\n",
    "y =[]\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [],
   "source": [
    "for features, label in training_data:\n",
    "    X.append(features)\n",
    "    y.append(label)\n",
    "\n",
    "X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "pickle_out =open('X.pickle', \"wb\")\n",
    "pickle.dump(X, pickle_out)\n",
    "pickle_out.close()\n",
    "\n",
    "pickle_out = open(\"y.pickle\", \"wb\")\n",
    "pickle.dump(y, pickle_out)\n",
    "pickle_out.close()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle_in = open(\"X.pickle\", \"rb\")\n",
    "X = pickle.load(pickle_in)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[58],\n",
       "        [48],\n",
       "        [49],\n",
       "        ...,\n",
       "        [57],\n",
       "        [56],\n",
       "        [55]],\n",
       "\n",
       "       [[73],\n",
       "        [57],\n",
       "        [53],\n",
       "        ...,\n",
       "        [58],\n",
       "        [57],\n",
       "        [56]],\n",
       "\n",
       "       [[87],\n",
       "        [66],\n",
       "        [55],\n",
       "        ...,\n",
       "        [59],\n",
       "        [59],\n",
       "        [57]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[48],\n",
       "        [47],\n",
       "        [47],\n",
       "        ...,\n",
       "        [91],\n",
       "        [80],\n",
       "        [74]],\n",
       "\n",
       "       [[49],\n",
       "        [48],\n",
       "        [48],\n",
       "        ...,\n",
       "        [89],\n",
       "        [82],\n",
       "        [75]],\n",
       "\n",
       "       [[50],\n",
       "        [49],\n",
       "        [48],\n",
       "        ...,\n",
       "        [76],\n",
       "        [73],\n",
       "        [66]]], dtype=uint8)"
      ]
     },
     "execution_count": 271,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "5/5 [==============================] - ETA: 0s - loss: 4.1234 - accuracy: 0.54 - 0s 65ms/step - loss: 4.0567 - accuracy: 0.5547 - val_loss: 10.5520 - val_accuracy: 0.4565\n",
      "Epoch 2/10\n",
      "5/5 [==============================] - 0s 45ms/step - loss: 5.3152 - accuracy: 0.5766 - val_loss: 3.1807 - val_accuracy: 0.5435\n",
      "Epoch 3/10\n",
      "5/5 [==============================] - 0s 48ms/step - loss: 2.9117 - accuracy: 0.5182 - val_loss: 1.2763 - val_accuracy: 0.5000\n",
      "Epoch 4/10\n",
      "5/5 [==============================] - 0s 44ms/step - loss: 1.2613 - accuracy: 0.5620 - val_loss: 1.4762 - val_accuracy: 0.4783\n",
      "Epoch 5/10\n",
      "5/5 [==============================] - 0s 50ms/step - loss: 0.3894 - accuracy: 0.8321 - val_loss: 0.8487 - val_accuracy: 0.6087\n",
      "Epoch 6/10\n",
      "5/5 [==============================] - 0s 48ms/step - loss: 0.6025 - accuracy: 0.5693 - val_loss: 0.5299 - val_accuracy: 0.6304\n",
      "Epoch 7/10\n",
      "5/5 [==============================] - 0s 47ms/step - loss: 0.2596 - accuracy: 0.9708 - val_loss: 0.7105 - val_accuracy: 0.6304\n",
      "Epoch 8/10\n",
      "5/5 [==============================] - 0s 44ms/step - loss: 0.2857 - accuracy: 0.9197 - val_loss: 0.7143 - val_accuracy: 0.6522\n",
      "Epoch 9/10\n",
      "5/5 [==============================] - 0s 45ms/step - loss: 0.2200 - accuracy: 0.9635 - val_loss: 0.5426 - val_accuracy: 0.6739\n",
      "Epoch 10/10\n",
      "5/5 [==============================] - 0s 45ms/step - loss: 0.1898 - accuracy: 0.9708 - val_loss: 0.5239 - val_accuracy: 0.6739\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1bd844b9970>"
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# CREATING THE MODEL WITH OUR LOADED DATA\n",
    "\n",
    "#for future reference...\n",
    "# A nice alternative to pickle is\n",
    "# np.save('features.npy',X) #saving\n",
    "# X=np.load('features.npy')#loading\n",
    "\n",
    "import tensorflow as tf\n",
    "import pickle\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Flatten, Conv2D, Activation, MaxPooling2D, Dropout\n",
    "\n",
    "\n",
    "X = pickle.load(open(\"X.pickle\",\"rb\"))\n",
    "y = pickle.load(open(\"y.pickle\",\"rb\"))\n",
    "\n",
    "X = np.array(X/255.0)\n",
    "y = np.array(y)\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Conv2D(64, (3,3), input_shape = X.shape[1:]))\n",
    "model.add(Activation(\"relu\"))\n",
    "model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "\n",
    "# model.add(Conv2D(64, (3,3))\n",
    "# model.add(Activation(\"relu\"))\n",
    "# model.add(MaxPooling2D(pool_size=(2,2)))\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(64))\n",
    "\n",
    "model.add(Dense(1))\n",
    "model.add(Activation('sigmoid'))\n",
    "          \n",
    "model.compile(loss=\"binary_crossentropy\",\n",
    "             optimizer=\"adam\",\n",
    "             metrics=['accuracy'])\n",
    "          \n",
    "model.fit(X, y, batch_size=32, epochs = 10, validation_split=0.25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6/6 [==============================] - 0s 12ms/step - loss: 0.2532 - accuracy: 0.9126\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.253150075674057, 0.9125683307647705]"
      ]
     },
     "execution_count": 274,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X, y)\n",
    "#[loss, accuracy]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: Eaglev1.model\\assets\n"
     ]
    }
   ],
   "source": [
    "model.save('Eaglev1.model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import tensorflow as tf\n",
    "\n",
    "CATEGORIES = [\"Eagle\", \"NonEagle\"]\n",
    "\n",
    "def prepare(filepath):\n",
    "    IMG_SIZE = 70  # 50 in txt-based\n",
    "    img_array = cv2.imread(filepath, cv2.IMREAD_GRAYSCALE)\n",
    "    new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))\n",
    "    return new_array.reshape(-1, IMG_SIZE, IMG_SIZE, 1)\n",
    "\n",
    "\n",
    "model = tf.keras.models.load_model(\"Eaglev1.model\")\n",
    "prediction = model.predict([prepare(r'C:\\Users\\chris\\Downloads\\EaglesMain\\vulture2.jpg')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.]]\n"
     ]
    }
   ],
   "source": [
    "print(prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NonEagle\n"
     ]
    }
   ],
   "source": [
    "\n",
    "print(CATEGORIES[int(prediction[0][0])] )\n",
    "#print(100 * predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#can continue retraining model with new inputs after loading\n",
    "#model.fit(newData, target, batch_size=32, epochs = 10, validation_split=0.25)\n",
    "#would have a file for collecting newly labeled data, and a trigger for the following;\n",
    "#will likely require having an additional data input, curate, train, and save\n",
    "#which would be followed by the statement of the new accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "11bd6e0ac4b74121946604cc6ddc3c16",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "FileUpload(value={}, description='Upload')"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "btn_upload = widgets.FileUpload()\n",
    "btn_upload"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 459,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0e6b1fca1caa4b628db9e0ef184f1a0a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "FileUpload(value={}, accept='.jpg', description='Upload')"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Creating widget U.I.\n",
    "\n",
    "#run interact for initiating retrain https://youtu.be/HY7cf-CYs1o?t=876\n",
    "import ipywidgets as widgets\n",
    "\n",
    "\n",
    "myupload = widgets.FileUpload(\n",
    "    accept='.jpg',  # Accepted file extension e.g. '.txt', '.pdf', 'image/*', 'image/*,.pdf'\n",
    "    multiple=False  # True to accept multiple files upload else False\n",
    ")\n",
    "myupload"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 389,
   "metadata": {},
   "outputs": [],
   "source": [
    "#myupload.value\n",
    "# uploaded_filename = next(iter(myupload.value))\n",
    "# content = myupload.value[uploaded_filename]['content']\n",
    "# with open('myfile', 'wb') as f: f.write(content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 397,
   "metadata": {},
   "outputs": [],
   "source": [
    "# myupload.value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 555,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install -Uqq fastbook\n",
    "import fastbook\n",
    "fastbook.setup_book()\n",
    "from fastbook import *\n",
    "from fastai.vision.widgets import *\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 556,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7db7c2286e4e4ee088e945a8cfd75337",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "FileUpload(value={}, description='Upload')"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "btn_upload = widgets.FileUpload()\n",
    "btn_upload\n",
    "# img = PILImage.create(btn_upload.data)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 557,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "list index out of range",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-557-075978508f7f>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mimg\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mPILImage\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbtn_upload\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mout_pl\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mwidgets\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mOutput\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[0mout_pl\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mclear_output\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;32mwith\u001b[0m \u001b[0mout_pl\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mdisplay\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto_thumb\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m70\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m70\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mout_pl\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mIndexError\u001b[0m: list index out of range"
     ]
    }
   ],
   "source": [
    "img = PILImage.create(btn_upload.data[-1])\n",
    "out_pl = widgets.Output()\n",
    "out_pl.clear_output()\n",
    "with out_pl: display(img.to_thumb(70,70))\n",
    "out_pl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 558,
   "metadata": {},
   "outputs": [],
   "source": [
    "# prediction = model.predict(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 559,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #test\n",
    "# from IPython.display import display\n",
    "# import FileUpload\n",
    "\n",
    "# uploader = fileupload.FileUploadWidget()\n",
    "\n",
    "# def _handle_upload(change):\n",
    "#     w = change['owner']\n",
    "#     with open(w.filename, 'wb') as f:\n",
    "#         f.write(w.data)\n",
    "#     print('Uploaded `{}` ({:.2f} kB)'.format(\n",
    "#         w.filename, len(w.data) / 2**10))\n",
    "\n",
    "# uploader.observe(_handle_upload, names='data')\n",
    "\n",
    "# display(uploader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 560,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "68aeb6c3d77a48cf881a8d9e6287ca21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "FileUpload(value={}, description='Upload')"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ipywidgets import FileUpload\n",
    "\n",
    "def on_upload_change(change):\n",
    "    if not change.new:\n",
    "        return\n",
    "    up = change.owner\n",
    "    for filename,data in up.value.items():\n",
    "        print(f'{filename}')\n",
    "        with open(filename, 'wb') as f:\n",
    "            f.write(data['content'])\n",
    "    up.value.clear()\n",
    "    up._counter = 0\n",
    "\n",
    "upload_btn = FileUpload()\n",
    "upload_btn.observe(on_upload_change, names='_counter')\n",
    "upload_btn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 561,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from ipywidgets import FileUpload\n",
    "\n",
    "# def on_upload_change(change):\n",
    "#     if not change.new:\n",
    "#         return\n",
    "#     up = change.owner\n",
    "#     for filename,data in up.value.items():\n",
    "#         print(f'writing [{filename}] to /Uploadsv1')\n",
    "#         with open(filename, 'wb') as f:\n",
    "#             f.write(data['content'])\n",
    "#             temp = f\n",
    "#     up.value.clear()\n",
    "#     up._counter = 0\n",
    "\n",
    "# upload_btn = FileUpload()\n",
    "# upload_btn.observe(on_upload_change, names='_counter')\n",
    "# upload_btn\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 548,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "list index out of range",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-548-8d8d2f8e5f93>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      2\u001b[0m \u001b[1;31m# btn_upload\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 4\u001b[1;33m \u001b[0mimg\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mPILImage\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mupload_btn\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      5\u001b[0m \u001b[0mout_pl\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mwidgets\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mOutput\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      6\u001b[0m \u001b[0mout_pl\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mclear_output\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mIndexError\u001b[0m: list index out of range"
     ]
    }
   ],
   "source": [
    "# btn_upload = widgets.FileUpload()\n",
    "# btn_upload\n",
    "\n",
    "img = PILImage.create(upload_btn.data[-1])\n",
    "out_pl = widgets.Output()\n",
    "out_pl.clear_output()\n",
    "with out_pl: display(img.to_thumb(70,70))\n",
    "out_pl\n",
    "\n",
    "#Should Display uploaded image; working"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 550,
   "metadata": {},
   "outputs": [
    {
     "ename": "error",
     "evalue": "OpenCV(4.5.1) C:\\Users\\appveyor\\AppData\\Local\\Temp\\1\\pip-req-build-kh7iq4w7\\opencv\\modules\\imgproc\\src\\resize.cpp:4051: error: (-215:Assertion failed) !ssize.empty() in function 'cv::resize'\n",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31merror\u001b[0m                                     Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-550-0ea704db4223>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mprediction\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mprepare\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtemp\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mCATEGORIES\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mprediction\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;31m#two main issues;\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;31m# 1 : all of them save to the current directory; not what I want\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[1;31m# 2 : I have to rename the variable manually after upload to get a prediction\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-334-dc2b69a0e1f9>\u001b[0m in \u001b[0;36mprepare\u001b[1;34m(filepath)\u001b[0m\n\u001b[0;32m      7\u001b[0m     \u001b[0mIMG_SIZE\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;36m70\u001b[0m  \u001b[1;31m# 50 in txt-based\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      8\u001b[0m     \u001b[0mimg_array\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcv2\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mimread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfilepath\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcv2\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mIMREAD_GRAYSCALE\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 9\u001b[1;33m     \u001b[0mnew_array\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcv2\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mresize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg_array\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mIMG_SIZE\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mIMG_SIZE\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     10\u001b[0m     \u001b[1;32mreturn\u001b[0m \u001b[0mnew_array\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mIMG_SIZE\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mIMG_SIZE\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31merror\u001b[0m: OpenCV(4.5.1) C:\\Users\\appveyor\\AppData\\Local\\Temp\\1\\pip-req-build-kh7iq4w7\\opencv\\modules\\imgproc\\src\\resize.cpp:4051: error: (-215:Assertion failed) !ssize.empty() in function 'cv::resize'\n"
     ]
    }
   ],
   "source": [
    "prediction = model.predict([prepare(###)])\n",
    "print(CATEGORIES[int(prediction[0][0])] )\n",
    "#two main issues;\n",
    "# 1 : all of them save to the current directory; not what I want\n",
    "# 2 : I have to rename the variable manually after upload to get a prediction\n",
    "\n",
    "#If I fix 2 I can move to setting up an initial binder and voila"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}