Pytorch ANN Exercise.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a2ba268a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.utils import shuffle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "baf76cc3",
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "44764a69",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"./income.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4541e51a",
   "metadata": {},
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>education</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>workclass</th>\n",
       "      <th>occupation</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>income</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>27</td>\n",
       "      <td>Male</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Private</td>\n",
       "      <td>Craft-repair</td>\n",
       "      <td>40</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>47</td>\n",
       "      <td>Male</td>\n",
       "      <td>Masters</td>\n",
       "      <td>14</td>\n",
       "      <td>Married</td>\n",
       "      <td>Local-gov</td>\n",
       "      <td>Exec-managerial</td>\n",
       "      <td>50</td>\n",
       "      <td>&gt;50K</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>59</td>\n",
       "      <td>Male</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Self-emp</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>20</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>38</td>\n",
       "      <td>Female</td>\n",
       "      <td>Prof-school</td>\n",
       "      <td>15</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Federal-gov</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>57</td>\n",
       "      <td>&gt;50K</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>64</td>\n",
       "      <td>Female</td>\n",
       "      <td>11th</td>\n",
       "      <td>7</td>\n",
       "      <td>Widowed</td>\n",
       "      <td>Private</td>\n",
       "      <td>Farming-fishing</td>\n",
       "      <td>40</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age     sex    education  education-num marital-status    workclass  \\\n",
       "0   27    Male      HS-grad              9  Never-married      Private   \n",
       "1   47    Male      Masters             14        Married    Local-gov   \n",
       "2   59    Male      HS-grad              9       Divorced     Self-emp   \n",
       "3   38  Female  Prof-school             15  Never-married  Federal-gov   \n",
       "4   64  Female         11th              7        Widowed      Private   \n",
       "\n",
       "        occupation  hours-per-week income  label  \n",
       "0     Craft-repair              40  <=50K      0  \n",
       "1  Exec-managerial              50   >50K      1  \n",
       "2   Prof-specialty              20  <=50K      0  \n",
       "3   Prof-specialty              57   >50K      1  \n",
       "4  Farming-fishing              40  <=50K      0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8d776294",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "label\n",
       "0    21700\n",
       "1     8300\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['label'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ca0f9612",
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_cols = ['sex', 'education', 'marital-status', 'workclass', 'occupation']\n",
    "cont_cols = ['age', 'hours-per-week']\n",
    "y_col = ['label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "33c2c03c",
   "metadata": {},
   "outputs": [],
   "source": [
    "for cat in cat_cols:\n",
    "    df[cat] = df[cat].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6d61728f",
   "metadata": {},
   "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>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>education</th>\n",
       "      <th>education-num</th>\n",
       "      <th>marital-status</th>\n",
       "      <th>workclass</th>\n",
       "      <th>occupation</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>income</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>23</td>\n",
       "      <td>Female</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Private</td>\n",
       "      <td>Other-service</td>\n",
       "      <td>50</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>37</td>\n",
       "      <td>Female</td>\n",
       "      <td>Prof-school</td>\n",
       "      <td>15</td>\n",
       "      <td>Married</td>\n",
       "      <td>State-gov</td>\n",
       "      <td>Prof-specialty</td>\n",
       "      <td>39</td>\n",
       "      <td>&gt;50K</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>34</td>\n",
       "      <td>Male</td>\n",
       "      <td>Some-college</td>\n",
       "      <td>10</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Private</td>\n",
       "      <td>Adm-clerical</td>\n",
       "      <td>40</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>31</td>\n",
       "      <td>Male</td>\n",
       "      <td>HS-grad</td>\n",
       "      <td>9</td>\n",
       "      <td>Married</td>\n",
       "      <td>Private</td>\n",
       "      <td>Craft-repair</td>\n",
       "      <td>40</td>\n",
       "      <td>&gt;50K</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>20</td>\n",
       "      <td>Female</td>\n",
       "      <td>Some-college</td>\n",
       "      <td>10</td>\n",
       "      <td>Never-married</td>\n",
       "      <td>Private</td>\n",
       "      <td>Sales</td>\n",
       "      <td>25</td>\n",
       "      <td>&lt;=50K</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age     sex     education  education-num marital-status  workclass  \\\n",
       "0   23  Female       HS-grad              9  Never-married    Private   \n",
       "1   37  Female   Prof-school             15        Married  State-gov   \n",
       "2   34    Male  Some-college             10       Divorced    Private   \n",
       "3   31    Male       HS-grad              9        Married    Private   \n",
       "4   20  Female  Some-college             10  Never-married    Private   \n",
       "\n",
       "       occupation  hours-per-week income  label  \n",
       "0   Other-service              50  <=50K      0  \n",
       "1  Prof-specialty              39   >50K      1  \n",
       "2    Adm-clerical              40  <=50K      0  \n",
       "3    Craft-repair              40   >50K      1  \n",
       "4           Sales              25  <=50K      0  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = shuffle(df, random_state=101)\n",
    "df.reset_index(drop=True, inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6f23145f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(2, 1), (14, 7), (6, 3), (5, 3), (12, 6)]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_szs = [len(df[col].cat.categories) for col in cat_cols]\n",
    "emb_szs = [(size, min(50, (size+1)//2)) for size in cat_szs]\n",
    "emb_szs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "489f739a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0, 10,  3,  2,  6],\n",
       "       [ 0, 12,  1,  4,  7],\n",
       "       [ 1, 13,  0,  2,  0],\n",
       "       [ 1, 10,  1,  2,  1],\n",
       "       [ 0, 13,  3,  2,  9]], dtype=int8)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "sx = df['sex'].cat.codes.values\n",
    "ed = df['education'].cat.codes.values\n",
    "ms = df['marital-status'].cat.codes.values\n",
    "wc = df['workclass'].cat.codes.values\n",
    "oc = df['occupation'].cat.codes.values\n",
    "\n",
    "cats = np.stack([sx,ed,ms,wc,oc], 1)\n",
    "\n",
    "cats[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c0974334",
   "metadata": {},
   "outputs": [],
   "source": [
    "cats = torch.tensor(cats, dtype=torch.int64)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c8a049aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[23, 50],\n",
       "       [37, 39],\n",
       "       [34, 40],\n",
       "       [31, 40],\n",
       "       [20, 25]])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conts = np.stack([df[col].values for col in cont_cols], 1)\n",
    "conts[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f28ffd96",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.float32"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conts = torch.tensor(conts, dtype=torch.float)\n",
    "conts.dtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "61283d10",
   "metadata": {},
   "outputs": [],
   "source": [
    "y = torch.tensor(df[y_col].values).flatten()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "97084814",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([0, 1, 0,  ..., 1, 0, 1])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "4e5e0c6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "b = 30000 # suggested batch size\n",
    "t = 5000  # suggested test size\n",
    "\n",
    "cat_train = cats[:b-t]\n",
    "cat_test  = cats[b-t:b]\n",
    "con_train = conts[:b-t]\n",
    "con_test  = conts[b-t:b]\n",
    "y_train   = y[:b-t]\n",
    "y_test    = y[b-t:b]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1f14cacb",
   "metadata": {},
   "outputs": [],
   "source": [
    "class TabularModel(nn.Module):\n",
    "\n",
    "    def __init__(self, emb_szs, n_cont, out_sz, layers, p=0.5):\n",
    "        # Call the parent __init__\n",
    "        super().__init__()\n",
    "        \n",
    "        # Set up the embedding, dropout, and batch normalization layer attributes\n",
    "        self.embeds = nn.ModuleList([nn.Embedding(ni, nf) for ni,nf in emb_szs])\n",
    "        self.emb_drop = nn.Dropout(p)\n",
    "        self.bn_cont = nn.BatchNorm1d(n_cont)\n",
    "        \n",
    "        # Assign a variable to hold a list of layers\n",
    "        layerlist = []\n",
    "        \n",
    "        # Assign a variable to store the number of embedding and continuous layers\n",
    "        n_emb = sum((nf for ni,nf in emb_szs))\n",
    "        n_in = n_emb + n_cont\n",
    "        \n",
    "        # Iterate through the passed-in \"layers\" parameter (ie, [200,100]) to build a list of layers\n",
    "        for i in layers:\n",
    "            layerlist.append(nn.Linear(n_in,i)) \n",
    "            layerlist.append(nn.ReLU(inplace=True))\n",
    "            layerlist.append(nn.BatchNorm1d(i))\n",
    "            layerlist.append(nn.Dropout(p))\n",
    "            n_in = i\n",
    "        layerlist.append(nn.Linear(layers[-1],out_sz))\n",
    "        \n",
    "        # Convert the list of layers into an attribute\n",
    "        self.layers = nn.Sequential(*layerlist)\n",
    "    \n",
    "    def forward(self, x_cat, x_cont):\n",
    "        # Extract embedding values from the incoming categorical data\n",
    "        embeddings = []\n",
    "        for i,e in enumerate(self.embeds):\n",
    "            embeddings.append(e(x_cat[:,i]))\n",
    "        x = torch.cat(embeddings, 1)\n",
    "        # Perform an initial dropout on the embeddings\n",
    "        x = self.emb_drop(x)\n",
    "        \n",
    "        # Normalize the incoming continuous data\n",
    "        x_cont = self.bn_cont(x_cont)\n",
    "        x = torch.cat([x, x_cont], 1)\n",
    "        \n",
    "        # Set up model layers\n",
    "        x = self.layers(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "fbd302c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7fa100f59610>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.manual_seed(33)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e8134ad9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TabularModel(\n",
       "  (embeds): ModuleList(\n",
       "    (0): Embedding(2, 1)\n",
       "    (1): Embedding(14, 7)\n",
       "    (2): Embedding(6, 3)\n",
       "    (3): Embedding(5, 3)\n",
       "    (4): Embedding(12, 6)\n",
       "  )\n",
       "  (emb_drop): Dropout(p=0.4, inplace=False)\n",
       "  (bn_cont): BatchNorm1d(2, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "  (layers): Sequential(\n",
       "    (0): Linear(in_features=22, out_features=50, bias=True)\n",
       "    (1): ReLU(inplace=True)\n",
       "    (2): BatchNorm1d(50, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (3): Dropout(p=0.4, inplace=False)\n",
       "    (4): Linear(in_features=50, out_features=2, bias=True)\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = TabularModel(emb_szs, conts.shape[1], 2, [50], p=0.4)\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "fe429703",
   "metadata": {},
   "outputs": [],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=0.001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "5ca6f6a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch:   1  loss: 0.31324983\n",
      "epoch:  26  loss: 0.31053504\n",
      "epoch:  51  loss: 0.30555499\n",
      "epoch:  76  loss: 0.30273521\n",
      "epoch: 101  loss: 0.30086908\n",
      "epoch: 126  loss: 0.29858872\n",
      "epoch: 151  loss: 0.29439232\n",
      "epoch: 176  loss: 0.29116037\n",
      "epoch: 201  loss: 0.28944671\n",
      "epoch: 226  loss: 0.28769010\n",
      "epoch: 251  loss: 0.28790537\n",
      "epoch: 276  loss: 0.28995341\n",
      "epoch: 300  loss: 0.29113221\n",
      "\n",
      "Duration: 28 seconds\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start_time = time.time()\n",
    "\n",
    "epochs = 300\n",
    "losses = []\n",
    "\n",
    "for i in range(epochs):\n",
    "    i+=1\n",
    "    y_pred = model(cat_train, con_train)\n",
    "    loss = criterion(y_pred, y_train)\n",
    "    losses.append(loss.item())\n",
    "    \n",
    "    # a neat trick to save screen space:\n",
    "    if i%25 == 1:\n",
    "        print(f'epoch: {i:3}  loss: {loss.item():10.8f}')\n",
    "\n",
    "    optimizer.zero_grad()\n",
    "    loss.backward()\n",
    "    optimizer.step()\n",
    "\n",
    "print(f'epoch: {i:3}  loss: {loss.item():10.8f}') # print the last line\n",
    "print(f'\\nDuration: {time.time() - start_time:.0f} seconds')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "617522b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(epochs), losses)\n",
    "plt.ylabel('Cross Entropy Loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "ee0abb98",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CE Loss: 0.31019413\n"
     ]
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    y_val = model(cat_test, con_test)\n",
    "    loss = criterion(y_val, y_test)\n",
    "print(f'CE Loss: {loss:.8f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "a13b7475",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "4272 out of 5000 = 85.44% correct\n"
     ]
    }
   ],
   "source": [
    "rows = len(y_test)\n",
    "correct = 0\n",
    "\n",
    "# print(f'{\"MODEL OUTPUT\":26} ARGMAX  Y_TEST')\n",
    "\n",
    "for i in range(rows):\n",
    "    # print(f'{str(y_val[i]):26} {y_val[i].argmax().item():^7}{y_test[i]:^7}')\n",
    "\n",
    "    if y_val[i].argmax().item() == y_test[i]:\n",
    "        correct += 1\n",
    "\n",
    "print(f'\\n{correct} out of {rows} = {100*correct/rows:.2f}% correct')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b563ae17",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "4272 out of 5000 = 85.44% correct\n"
     ]
    }
   ],
   "source": [
    "rows = len(y_test)\n",
    "correct = 0\n",
    "\n",
    "# print(f'{\"MODEL OUTPUT\":26} ARGMAX  Y_TEST')\n",
    "\n",
    "for i in range(rows):\n",
    "    # print(f'{str(y_val[i]):26} {y_val[i].argmax().item():^7}{y_test[i]:^7}')\n",
    "\n",
    "    if y_val[i].argmax().item() == y_test[i]:\n",
    "        correct += 1\n",
    "\n",
    "print(f'\\n{correct} out of {rows} = {100*correct/rows:.2f}% correct')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "785ecde7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "What is the person's age? (18-90)  22\n",
      "What is the person's sex? (Male/Female) Male\n",
      "What is the person's education level? (3-16) 4\n",
      "What is the person's marital status? Married\n",
      "What is the person's workclass? Federal-gov\n",
      "What is the person's occupation? Adm-clerical\n",
      "How many hours/week are worked? (20-90)  49\n",
      "\n",
      "The predicted label is 0\n"
     ]
    }
   ],
   "source": [
    "def test_data(mdl): # pass in the name of the model\n",
    "    # INPUT NEW DATA\n",
    "    age = float(input(\"What is the person's age? (18-90)  \"))\n",
    "    sex = input(\"What is the person's sex? (Male/Female) \").capitalize()\n",
    "    edn = int(input(\"What is the person's education level? (3-16) \"))\n",
    "    mar = input(\"What is the person's marital status? \").capitalize()\n",
    "    wrk = input(\"What is the person's workclass? \").capitalize()\n",
    "    occ = input(\"What is the person's occupation? \").capitalize()\n",
    "    hrs = float(input(\"How many hours/week are worked? (20-90)  \"))\n",
    "\n",
    "    # PREPROCESS THE DATA\n",
    "    sex_d = {'Female':0, 'Male':1}\n",
    "    mar_d = {'Divorced':0, 'Married':1, 'Married-spouse-absent':2, 'Never-married':3, 'Separated':4, 'Widowed':5}\n",
    "    wrk_d = {'Federal-gov':0, 'Local-gov':1, 'Private':2, 'Self-emp':3, 'State-gov':4}\n",
    "    occ_d = {'Adm-clerical':0, 'Craft-repair':1, 'Exec-managerial':2, 'Farming-fishing':3, 'Handlers-cleaners':4,\n",
    "            'Machine-op-inspct':5, 'Other-service':6, 'Prof-specialty':7, 'Protective-serv':8, 'Sales':9, \n",
    "            'Tech-support':10, 'Transport-moving':11}\n",
    "\n",
    "    sex = sex_d[sex]\n",
    "    mar = mar_d[mar]\n",
    "    wrk = wrk_d[wrk]\n",
    "    occ = occ_d[occ]\n",
    "\n",
    "    # CREATE CAT AND CONT TENSORS\n",
    "    cats = torch.tensor([sex,edn,mar,wrk,occ], dtype=torch.int64).reshape(1,-1)\n",
    "    conts = torch.tensor([age,hrs], dtype=torch.float).reshape(1,-1)\n",
    "    \n",
    "    # SET MODEL TO EVAL (in case this hasn't been done)\n",
    "    mdl.eval()\n",
    "\n",
    "    # PASS NEW DATA THROUGH THE MODEL WITHOUT PERFORMING A BACKPROP\n",
    "    with torch.no_grad():\n",
    "        z = mdl(cats, conts).argmax().item()\n",
    "\n",
    "    print(f'\\nThe predicted label is {z}')\n",
    "    \n",
    "test_data(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4422f0f0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "66159fd5",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a1fc406b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56454aea",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5361c039",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}