request_recommender_service.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests\n",
    "import numpy as np\n",
    "import json\n",
    "import time\n",
    "import pandas as pd\n",
    "from IPython.core.interactiveshell import InteractiveShell\n",
    "InteractiveShell.ast_node_interactivity = \"all\"\n",
    "\n",
    "np.set_printoptions(precision=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "NAMESPACE = 'kfserving-test'\n",
    "MODEL_NAME =  'product-recommender' #'autoencoder-recommender' 'product-recommender' 'concept-drift'\n",
    "KFSERVING_VERSION = 'v1beta1'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load preprocessed Data from Training of the Recommender Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(45916, 207, 52)\n",
      "(45916,)\n"
     ]
    },
    {
     "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>Unnamed: 0</th>\n",
       "      <th>Item_ID</th>\n",
       "      <th>Mapped_ID</th>\n",
       "      <th>category_code</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2402273</td>\n",
       "      <td>1</td>\n",
       "      <td>appliances.personal.massager</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>20100164</td>\n",
       "      <td>2</td>\n",
       "      <td>apparel.trousers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>21400264</td>\n",
       "      <td>3</td>\n",
       "      <td>electronics.clocks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1005239</td>\n",
       "      <td>4</td>\n",
       "      <td>construction.tools.light</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>5100885</td>\n",
       "      <td>5</td>\n",
       "      <td>computers.notebook</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38510</th>\n",
       "      <td>38510</td>\n",
       "      <td>6902925</td>\n",
       "      <td>38511</td>\n",
       "      <td>electronics.telephone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38511</th>\n",
       "      <td>38511</td>\n",
       "      <td>2602249</td>\n",
       "      <td>38512</td>\n",
       "      <td>appliances.kitchen.refrigerators</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38512</th>\n",
       "      <td>38512</td>\n",
       "      <td>20100099</td>\n",
       "      <td>38513</td>\n",
       "      <td>apparel.trousers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38513</th>\n",
       "      <td>38513</td>\n",
       "      <td>21405267</td>\n",
       "      <td>38514</td>\n",
       "      <td>electronics.clocks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38514</th>\n",
       "      <td>38514</td>\n",
       "      <td>26405181</td>\n",
       "      <td>38515</td>\n",
       "      <td>computers.peripherals.printer</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>38515 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0   Item_ID  Mapped_ID                     category_code\n",
       "0               0   2402273          1      appliances.personal.massager\n",
       "1               1  20100164          2                  apparel.trousers\n",
       "2               2  21400264          3                electronics.clocks\n",
       "3               3   1005239          4          construction.tools.light\n",
       "4               4   5100885          5                computers.notebook\n",
       "...           ...       ...        ...                               ...\n",
       "38510       38510   6902925      38511             electronics.telephone\n",
       "38511       38511   2602249      38512  appliances.kitchen.refrigerators\n",
       "38512       38512  20100099      38513                  apparel.trousers\n",
       "38513       38513  21405267      38514                electronics.clocks\n",
       "38514       38514  26405181      38515     computers.peripherals.printer\n",
       "\n",
       "[38515 rows x 4 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sessions_padded = np.load('list_sessions_padded.npy')\n",
    "print(sessions_padded.shape)\n",
    "last_clicked = np.load('list_last_clicked.npy')\n",
    "print(last_clicked.shape)\n",
    "n_output_features = int(last_clicked.max())\n",
    "n_unique_input_ids = int(sessions_padded[:,:,0].max())\n",
    "window_length = sessions_padded.shape[1]\n",
    "n_input_features = sessions_padded.shape[2]\n",
    "\n",
    "id_mapping = pd.read_csv('ID_Mapping.csv')\n",
    "id_mapping"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Request Recommender Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.1285841464996338\n",
      "Session:\n"
     ]
    },
    {
     "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>category_code</th>\n",
       "      <th>Item_ID</th>\n",
       "      <th>Item_ID_Mapped</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>kids.skates</td>\n",
       "      <td>23301564</td>\n",
       "      <td>10804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>construction.tools.generator</td>\n",
       "      <td>15901606</td>\n",
       "      <td>2963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>furniture.kitchen.table</td>\n",
       "      <td>31300060</td>\n",
       "      <td>24062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>apparel.shoes.keds</td>\n",
       "      <td>100011061</td>\n",
       "      <td>559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>apparel.shoes.keds</td>\n",
       "      <td>16400390</td>\n",
       "      <td>2757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>apparel.shoes.keds</td>\n",
       "      <td>100011061</td>\n",
       "      <td>559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>auto.accessories.winch</td>\n",
       "      <td>24400437</td>\n",
       "      <td>24615</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  category_code    Item_ID  Item_ID_Mapped\n",
       "0                   kids.skates   23301564           10804\n",
       "1  construction.tools.generator   15901606            2963\n",
       "2       furniture.kitchen.table   31300060           24062\n",
       "3            apparel.shoes.keds  100011061             559\n",
       "4            apparel.shoes.keds   16400390            2757\n",
       "5            apparel.shoes.keds  100011061             559\n",
       "6        auto.accessories.winch   24400437           24615"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction:\n"
     ]
    },
    {
     "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>category_code</th>\n",
       "      <th>Item_ID</th>\n",
       "      <th>Item_ID_Mapped</th>\n",
       "      <th>probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>apparel.shoes</td>\n",
       "      <td>5300300</td>\n",
       "      <td>24969</td>\n",
       "      <td>0.999923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>appliances.kitchen.kettle</td>\n",
       "      <td>20000602</td>\n",
       "      <td>17781</td>\n",
       "      <td>0.000022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>electronics.clocks</td>\n",
       "      <td>5100266</td>\n",
       "      <td>14029</td>\n",
       "      <td>0.000006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>electronics.clocks</td>\n",
       "      <td>100005669</td>\n",
       "      <td>1928</td>\n",
       "      <td>0.000005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>electronics.video.tv</td>\n",
       "      <td>1802075</td>\n",
       "      <td>21568</td>\n",
       "      <td>0.000004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               category_code    Item_ID  Item_ID_Mapped  probability\n",
       "0              apparel.shoes    5300300           24969     0.999923\n",
       "1  appliances.kitchen.kettle   20000602           17781     0.000022\n",
       "2         electronics.clocks    5100266           14029     0.000006\n",
       "3         electronics.clocks  100005669            1928     0.000005\n",
       "4       electronics.video.tv    1802075           21568     0.000004"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ground Truth: 24969\n"
     ]
    }
   ],
   "source": [
    "def request_kf_serving(np_array, MODEL_NAME, NAMESPACE, INGRESS_HOST, INGRESS_PORT, item_ground_truth):\n",
    "    data = json.dumps({\"instances\": np_array.tolist(),\n",
    "                      'id': item_ground_truth})\n",
    "    headers = {#\"content-type\": \"application/json\",\n",
    "               'Host': '{}.{}.example.com'.format(MODEL_NAME, NAMESPACE)}\n",
    "    json_response = requests.post(\n",
    "        'http://{}:{}/v1/models/{}:predict'.format(INGRESS_HOST, INGRESS_PORT, MODEL_NAME), data=data, headers=headers)\n",
    "\n",
    "    try:\n",
    "        predictions = json.loads(json_response.text)['predictions']\n",
    "    except Exception as e:\n",
    "        raise e\n",
    "    return np.array(predictions).astype(np.float32)\n",
    "\n",
    "\n",
    "\n",
    "INGRESS_HOST_LIST = !kubectl get po -l istio=ingressgateway -n istio-system -o jsonpath={.items[0].status.hostIP}\n",
    "INGRESS_HOST =  INGRESS_HOST_LIST[0] #eg. '192.168.52.86'\n",
    "#print(INGRESS_HOST)\n",
    "INGRESS_PORT = 31017\n",
    "\n",
    "example = 14 #5,6,7, 13\n",
    "start = time.time()\n",
    "\n",
    "pred = request_kf_serving(sessions_padded[example:example+1],#[np.newaxis,:,:],\n",
    "                          MODEL_NAME, NAMESPACE, INGRESS_HOST, INGRESS_PORT, int(last_clicked[example]))   \n",
    "end = time.time()\n",
    "print(end - start)\n",
    "\n",
    "top = pred.argsort()[0][::-1][:5]\n",
    "print(\"Session:\")\n",
    "session = pd.DataFrame()\n",
    "session['category_code'] = [id_mapping['category_code'][int(i)-1] for i in sessions_padded[example,:,0] if i>0]\n",
    "session['Item_ID'] = [id_mapping['Item_ID'][int(i)-1] for i in sessions_padded[example,:,0] if i>0]\n",
    "session['Item_ID_Mapped'] = [int(i) for i in sessions_padded[example,:,0] if i>0]\n",
    "session\n",
    "\n",
    "\n",
    "print(\"Prediction:\")\n",
    "prediction = pd.DataFrame()\n",
    "prediction['category_code'] = [id_mapping['category_code'][int(i)-1] for i in top if i>0]\n",
    "prediction['Item_ID'] = [id_mapping['Item_ID'][int(i)-1] for i in top if i>0]\n",
    "prediction['Item_ID_Mapped'] = [int(i) for i in top if i>0]\n",
    "prediction['probability'] = pred[0, top]\n",
    "prediction\n",
    "print(\"Ground Truth:\", last_clicked[example])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test autoscaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import threading\n",
    "\n",
    "while True:\n",
    "    for _ in range(10):\n",
    "        x = threading.Thread(target=request_kf_serving, args=(np.ones((1, 207, 52), dtype=np.float32),\n",
    "                               MODEL_NAME, NAMESPACE, INGRESS_HOST, INGRESS_PORT))\n",
    "        x.start()\n",
    "    time.sleep(0.3)"
   ]
  },
  {
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}