├── README.md
├── SalaryPrediction.ipynb
├── app.py
├── explore_page.py
└── predict_page.py


/README.md:
--------------------------------------------------------------------------------
1 | # Build A Salary Prediction Web App With Streamlit
2 | 
3 | Build a Machine Learning web application from scratch in Python with Streamlit. We use real world data to build a machine learning model. In the first part of the video you learn how we analyze the data and build our model, and in the second part we build the web app using streamlit.
4 | 
5 | Watch the video on YouTube:  
6 | [![Alt text](https://img.youtube.com/vi/xl0N7tHiwlw/hqdefault.jpg)](https://youtu.be/xl0N7tHiwlw)  
7 | [https://youtu.be/xl0N7tHiwlw](https://youtu.be/xl0N7tHiwlw)
8 | 


--------------------------------------------------------------------------------
/SalaryPrediction.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |  "cells": [
   3 |   {
   4 |    "cell_type": "code",
   5 |    "execution_count": 55,
   6 |    "id": "8b11fd46",
   7 |    "metadata": {},
   8 |    "outputs": [],
   9 |    "source": [
  10 |     "import pandas as pd\n",
  11 |     "import matplotlib.pyplot as plt\n",
  12 |     "\n",
  13 |     "df = pd.read_csv(\"survey_results_public.csv\")"
  14 |    ]
  15 |   },
  16 |   {
  17 |    "cell_type": "code",
  18 |    "execution_count": 56,
  19 |    "id": "0a94c801",
  20 |    "metadata": {},
  21 |    "outputs": [
  22 |     {
  23 |      "data": {
  24 |       "text/html": [
  25 |        "<div>\n",
  26 |        "<style scoped>\n",
  27 |        "    .dataframe tbody tr th:only-of-type {\n",
  28 |        "        vertical-align: middle;\n",
  29 |        "    }\n",
  30 |        "\n",
  31 |        "    .dataframe tbody tr th {\n",
  32 |        "        vertical-align: top;\n",
  33 |        "    }\n",
  34 |        "\n",
  35 |        "    .dataframe thead th {\n",
  36 |        "        text-align: right;\n",
  37 |        "    }\n",
  38 |        "</style>\n",
  39 |        "<table border=\"1\" class=\"dataframe\">\n",
  40 |        "  <thead>\n",
  41 |        "    <tr style=\"text-align: right;\">\n",
  42 |        "      <th></th>\n",
  43 |        "      <th>Respondent</th>\n",
  44 |        "      <th>MainBranch</th>\n",
  45 |        "      <th>Hobbyist</th>\n",
  46 |        "      <th>Age</th>\n",
  47 |        "      <th>Age1stCode</th>\n",
  48 |        "      <th>CompFreq</th>\n",
  49 |        "      <th>CompTotal</th>\n",
  50 |        "      <th>ConvertedComp</th>\n",
  51 |        "      <th>Country</th>\n",
  52 |        "      <th>CurrencyDesc</th>\n",
  53 |        "      <th>...</th>\n",
  54 |        "      <th>SurveyEase</th>\n",
  55 |        "      <th>SurveyLength</th>\n",
  56 |        "      <th>Trans</th>\n",
  57 |        "      <th>UndergradMajor</th>\n",
  58 |        "      <th>WebframeDesireNextYear</th>\n",
  59 |        "      <th>WebframeWorkedWith</th>\n",
  60 |        "      <th>WelcomeChange</th>\n",
  61 |        "      <th>WorkWeekHrs</th>\n",
  62 |        "      <th>YearsCode</th>\n",
  63 |        "      <th>YearsCodePro</th>\n",
  64 |        "    </tr>\n",
  65 |        "  </thead>\n",
  66 |        "  <tbody>\n",
  67 |        "    <tr>\n",
  68 |        "      <th>0</th>\n",
  69 |        "      <td>1</td>\n",
  70 |        "      <td>I am a developer by profession</td>\n",
  71 |        "      <td>Yes</td>\n",
  72 |        "      <td>NaN</td>\n",
  73 |        "      <td>13</td>\n",
  74 |        "      <td>Monthly</td>\n",
  75 |        "      <td>NaN</td>\n",
  76 |        "      <td>NaN</td>\n",
  77 |        "      <td>Germany</td>\n",
  78 |        "      <td>European Euro</td>\n",
  79 |        "      <td>...</td>\n",
  80 |        "      <td>Neither easy nor difficult</td>\n",
  81 |        "      <td>Appropriate in length</td>\n",
  82 |        "      <td>No</td>\n",
  83 |        "      <td>Computer science, computer engineering, or sof...</td>\n",
  84 |        "      <td>ASP.NET Core</td>\n",
  85 |        "      <td>ASP.NET;ASP.NET Core</td>\n",
  86 |        "      <td>Just as welcome now as I felt last year</td>\n",
  87 |        "      <td>50.0</td>\n",
  88 |        "      <td>36</td>\n",
  89 |        "      <td>27</td>\n",
  90 |        "    </tr>\n",
  91 |        "    <tr>\n",
  92 |        "      <th>1</th>\n",
  93 |        "      <td>2</td>\n",
  94 |        "      <td>I am a developer by profession</td>\n",
  95 |        "      <td>No</td>\n",
  96 |        "      <td>NaN</td>\n",
  97 |        "      <td>19</td>\n",
  98 |        "      <td>NaN</td>\n",
  99 |        "      <td>NaN</td>\n",
 100 |        "      <td>NaN</td>\n",
 101 |        "      <td>United Kingdom</td>\n",
 102 |        "      <td>Pound sterling</td>\n",
 103 |        "      <td>...</td>\n",
 104 |        "      <td>NaN</td>\n",
 105 |        "      <td>NaN</td>\n",
 106 |        "      <td>NaN</td>\n",
 107 |        "      <td>Computer science, computer engineering, or sof...</td>\n",
 108 |        "      <td>NaN</td>\n",
 109 |        "      <td>NaN</td>\n",
 110 |        "      <td>Somewhat more welcome now than last year</td>\n",
 111 |        "      <td>NaN</td>\n",
 112 |        "      <td>7</td>\n",
 113 |        "      <td>4</td>\n",
 114 |        "    </tr>\n",
 115 |        "    <tr>\n",
 116 |        "      <th>2</th>\n",
 117 |        "      <td>3</td>\n",
 118 |        "      <td>I code primarily as a hobby</td>\n",
 119 |        "      <td>Yes</td>\n",
 120 |        "      <td>NaN</td>\n",
 121 |        "      <td>15</td>\n",
 122 |        "      <td>NaN</td>\n",
 123 |        "      <td>NaN</td>\n",
 124 |        "      <td>NaN</td>\n",
 125 |        "      <td>Russian Federation</td>\n",
 126 |        "      <td>NaN</td>\n",
 127 |        "      <td>...</td>\n",
 128 |        "      <td>Neither easy nor difficult</td>\n",
 129 |        "      <td>Appropriate in length</td>\n",
 130 |        "      <td>NaN</td>\n",
 131 |        "      <td>NaN</td>\n",
 132 |        "      <td>NaN</td>\n",
 133 |        "      <td>NaN</td>\n",
 134 |        "      <td>Somewhat more welcome now than last year</td>\n",
 135 |        "      <td>NaN</td>\n",
 136 |        "      <td>4</td>\n",
 137 |        "      <td>NaN</td>\n",
 138 |        "    </tr>\n",
 139 |        "    <tr>\n",
 140 |        "      <th>3</th>\n",
 141 |        "      <td>4</td>\n",
 142 |        "      <td>I am a developer by profession</td>\n",
 143 |        "      <td>Yes</td>\n",
 144 |        "      <td>25.0</td>\n",
 145 |        "      <td>18</td>\n",
 146 |        "      <td>NaN</td>\n",
 147 |        "      <td>NaN</td>\n",
 148 |        "      <td>NaN</td>\n",
 149 |        "      <td>Albania</td>\n",
 150 |        "      <td>Albanian lek</td>\n",
 151 |        "      <td>...</td>\n",
 152 |        "      <td>NaN</td>\n",
 153 |        "      <td>NaN</td>\n",
 154 |        "      <td>No</td>\n",
 155 |        "      <td>Computer science, computer engineering, or sof...</td>\n",
 156 |        "      <td>NaN</td>\n",
 157 |        "      <td>NaN</td>\n",
 158 |        "      <td>Somewhat less welcome now than last year</td>\n",
 159 |        "      <td>40.0</td>\n",
 160 |        "      <td>7</td>\n",
 161 |        "      <td>4</td>\n",
 162 |        "    </tr>\n",
 163 |        "    <tr>\n",
 164 |        "      <th>4</th>\n",
 165 |        "      <td>5</td>\n",
 166 |        "      <td>I used to be a developer by profession, but no...</td>\n",
 167 |        "      <td>Yes</td>\n",
 168 |        "      <td>31.0</td>\n",
 169 |        "      <td>16</td>\n",
 170 |        "      <td>NaN</td>\n",
 171 |        "      <td>NaN</td>\n",
 172 |        "      <td>NaN</td>\n",
 173 |        "      <td>United States</td>\n",
 174 |        "      <td>NaN</td>\n",
 175 |        "      <td>...</td>\n",
 176 |        "      <td>Easy</td>\n",
 177 |        "      <td>Too short</td>\n",
 178 |        "      <td>No</td>\n",
 179 |        "      <td>Computer science, computer engineering, or sof...</td>\n",
 180 |        "      <td>Django;Ruby on Rails</td>\n",
 181 |        "      <td>Ruby on Rails</td>\n",
 182 |        "      <td>Just as welcome now as I felt last year</td>\n",
 183 |        "      <td>NaN</td>\n",
 184 |        "      <td>15</td>\n",
 185 |        "      <td>8</td>\n",
 186 |        "    </tr>\n",
 187 |        "  </tbody>\n",
 188 |        "</table>\n",
 189 |        "<p>5 rows × 61 columns</p>\n",
 190 |        "</div>"
 191 |       ],
 192 |       "text/plain": [
 193 |        "   Respondent                                         MainBranch Hobbyist  \\\n",
 194 |        "0           1                     I am a developer by profession      Yes   \n",
 195 |        "1           2                     I am a developer by profession       No   \n",
 196 |        "2           3                        I code primarily as a hobby      Yes   \n",
 197 |        "3           4                     I am a developer by profession      Yes   \n",
 198 |        "4           5  I used to be a developer by profession, but no...      Yes   \n",
 199 |        "\n",
 200 |        "    Age Age1stCode CompFreq  CompTotal  ConvertedComp             Country  \\\n",
 201 |        "0   NaN         13  Monthly        NaN            NaN             Germany   \n",
 202 |        "1   NaN         19      NaN        NaN            NaN      United Kingdom   \n",
 203 |        "2   NaN         15      NaN        NaN            NaN  Russian Federation   \n",
 204 |        "3  25.0         18      NaN        NaN            NaN             Albania   \n",
 205 |        "4  31.0         16      NaN        NaN            NaN       United States   \n",
 206 |        "\n",
 207 |        "     CurrencyDesc  ...                  SurveyEase           SurveyLength  \\\n",
 208 |        "0   European Euro  ...  Neither easy nor difficult  Appropriate in length   \n",
 209 |        "1  Pound sterling  ...                         NaN                    NaN   \n",
 210 |        "2             NaN  ...  Neither easy nor difficult  Appropriate in length   \n",
 211 |        "3    Albanian lek  ...                         NaN                    NaN   \n",
 212 |        "4             NaN  ...                        Easy              Too short   \n",
 213 |        "\n",
 214 |        "  Trans                                     UndergradMajor  \\\n",
 215 |        "0    No  Computer science, computer engineering, or sof...   \n",
 216 |        "1   NaN  Computer science, computer engineering, or sof...   \n",
 217 |        "2   NaN                                                NaN   \n",
 218 |        "3    No  Computer science, computer engineering, or sof...   \n",
 219 |        "4    No  Computer science, computer engineering, or sof...   \n",
 220 |        "\n",
 221 |        "  WebframeDesireNextYear    WebframeWorkedWith  \\\n",
 222 |        "0           ASP.NET Core  ASP.NET;ASP.NET Core   \n",
 223 |        "1                    NaN                   NaN   \n",
 224 |        "2                    NaN                   NaN   \n",
 225 |        "3                    NaN                   NaN   \n",
 226 |        "4   Django;Ruby on Rails         Ruby on Rails   \n",
 227 |        "\n",
 228 |        "                              WelcomeChange WorkWeekHrs YearsCode YearsCodePro  \n",
 229 |        "0   Just as welcome now as I felt last year        50.0        36           27  \n",
 230 |        "1  Somewhat more welcome now than last year         NaN         7            4  \n",
 231 |        "2  Somewhat more welcome now than last year         NaN         4          NaN  \n",
 232 |        "3  Somewhat less welcome now than last year        40.0         7            4  \n",
 233 |        "4   Just as welcome now as I felt last year         NaN        15            8  \n",
 234 |        "\n",
 235 |        "[5 rows x 61 columns]"
 236 |       ]
 237 |      },
 238 |      "execution_count": 56,
 239 |      "metadata": {},
 240 |      "output_type": "execute_result"
 241 |     }
 242 |    ],
 243 |    "source": [
 244 |     "df.head()"
 245 |    ]
 246 |   },
 247 |   {
 248 |    "cell_type": "code",
 249 |    "execution_count": 57,
 250 |    "id": "2e754e05",
 251 |    "metadata": {},
 252 |    "outputs": [
 253 |     {
 254 |      "data": {
 255 |       "text/html": [
 256 |        "<div>\n",
 257 |        "<style scoped>\n",
 258 |        "    .dataframe tbody tr th:only-of-type {\n",
 259 |        "        vertical-align: middle;\n",
 260 |        "    }\n",
 261 |        "\n",
 262 |        "    .dataframe tbody tr th {\n",
 263 |        "        vertical-align: top;\n",
 264 |        "    }\n",
 265 |        "\n",
 266 |        "    .dataframe thead th {\n",
 267 |        "        text-align: right;\n",
 268 |        "    }\n",
 269 |        "</style>\n",
 270 |        "<table border=\"1\" class=\"dataframe\">\n",
 271 |        "  <thead>\n",
 272 |        "    <tr style=\"text-align: right;\">\n",
 273 |        "      <th></th>\n",
 274 |        "      <th>Country</th>\n",
 275 |        "      <th>EdLevel</th>\n",
 276 |        "      <th>YearsCodePro</th>\n",
 277 |        "      <th>Employment</th>\n",
 278 |        "      <th>Salary</th>\n",
 279 |        "    </tr>\n",
 280 |        "  </thead>\n",
 281 |        "  <tbody>\n",
 282 |        "    <tr>\n",
 283 |        "      <th>0</th>\n",
 284 |        "      <td>Germany</td>\n",
 285 |        "      <td>Master’s degree (M.A., M.S., M.Eng., MBA, etc.)</td>\n",
 286 |        "      <td>27</td>\n",
 287 |        "      <td>Independent contractor, freelancer, or self-em...</td>\n",
 288 |        "      <td>NaN</td>\n",
 289 |        "    </tr>\n",
 290 |        "    <tr>\n",
 291 |        "      <th>1</th>\n",
 292 |        "      <td>United Kingdom</td>\n",
 293 |        "      <td>Bachelor’s degree (B.A., B.S., B.Eng., etc.)</td>\n",
 294 |        "      <td>4</td>\n",
 295 |        "      <td>Employed full-time</td>\n",
 296 |        "      <td>NaN</td>\n",
 297 |        "    </tr>\n",
 298 |        "    <tr>\n",
 299 |        "      <th>2</th>\n",
 300 |        "      <td>Russian Federation</td>\n",
 301 |        "      <td>NaN</td>\n",
 302 |        "      <td>NaN</td>\n",
 303 |        "      <td>NaN</td>\n",
 304 |        "      <td>NaN</td>\n",
 305 |        "    </tr>\n",
 306 |        "    <tr>\n",
 307 |        "      <th>3</th>\n",
 308 |        "      <td>Albania</td>\n",
 309 |        "      <td>Master’s degree (M.A., M.S., M.Eng., MBA, etc.)</td>\n",
 310 |        "      <td>4</td>\n",
 311 |        "      <td>NaN</td>\n",
 312 |        "      <td>NaN</td>\n",
 313 |        "    </tr>\n",
 314 |        "    <tr>\n",
 315 |        "      <th>4</th>\n",
 316 |        "      <td>United States</td>\n",
 317 |        "      <td>Bachelor’s degree (B.A., B.S., B.Eng., etc.)</td>\n",
 318 |        "      <td>8</td>\n",
 319 |        "      <td>Employed full-time</td>\n",
 320 |        "      <td>NaN</td>\n",
 321 |        "    </tr>\n",
 322 |        "  </tbody>\n",
 323 |        "</table>\n",
 324 |        "</div>"
 325 |       ],
 326 |       "text/plain": [
 327 |        "              Country                                          EdLevel  \\\n",
 328 |        "0             Germany  Master’s degree (M.A., M.S., M.Eng., MBA, etc.)   \n",
 329 |        "1      United Kingdom     Bachelor’s degree (B.A., B.S., B.Eng., etc.)   \n",
 330 |        "2  Russian Federation                                              NaN   \n",
 331 |        "3             Albania  Master’s degree (M.A., M.S., M.Eng., MBA, etc.)   \n",
 332 |        "4       United States     Bachelor’s degree (B.A., B.S., B.Eng., etc.)   \n",
 333 |        "\n",
 334 |        "  YearsCodePro                                         Employment  Salary  \n",
 335 |        "0           27  Independent contractor, freelancer, or self-em...     NaN  \n",
 336 |        "1            4                                 Employed full-time     NaN  \n",
 337 |        "2          NaN                                                NaN     NaN  \n",
 338 |        "3            4                                                NaN     NaN  \n",
 339 |        "4            8                                 Employed full-time     NaN  "
 340 |       ]
 341 |      },
 342 |      "execution_count": 57,
 343 |      "metadata": {},
 344 |      "output_type": "execute_result"
 345 |     }
 346 |    ],
 347 |    "source": [
 348 |     "df = df[[\"Country\", \"EdLevel\", \"YearsCodePro\", \"Employment\", \"ConvertedComp\"]]\n",
 349 |     "df = df.rename({\"ConvertedComp\": \"Salary\"}, axis=1)\n",
 350 |     "df.head()"
 351 |    ]
 352 |   },
 353 |   {
 354 |    "cell_type": "code",
 355 |    "execution_count": 58,
 356 |    "id": "c7653efa",
 357 |    "metadata": {},
 358 |    "outputs": [
 359 |     {
 360 |      "data": {
 361 |       "text/html": [
 362 |        "<div>\n",
 363 |        "<style scoped>\n",
 364 |        "    .dataframe tbody tr th:only-of-type {\n",
 365 |        "        vertical-align: middle;\n",
 366 |        "    }\n",
 367 |        "\n",
 368 |        "    .dataframe tbody tr th {\n",
 369 |        "        vertical-align: top;\n",
 370 |        "    }\n",
 371 |        "\n",
 372 |        "    .dataframe thead th {\n",
 373 |        "        text-align: right;\n",
 374 |        "    }\n",
 375 |        "</style>\n",
 376 |        "<table border=\"1\" class=\"dataframe\">\n",
 377 |        "  <thead>\n",
 378 |        "    <tr style=\"text-align: right;\">\n",
 379 |        "      <th></th>\n",
 380 |        "      <th>Country</th>\n",
 381 |        "      <th>EdLevel</th>\n",
 382 |        "      <th>YearsCodePro</th>\n",
 383 |        "      <th>Employment</th>\n",
 384 |        "      <th>Salary</th>\n",
 385 |        "    </tr>\n",
 386 |        "  </thead>\n",
 387 |        "  <tbody>\n",
 388 |        "    <tr>\n",
 389 |        "      <th>7</th>\n",
 390 |        "      <td>United States</td>\n",
 391 |        "      <td>Bachelor’s degree (B.A., B.S., B.Eng., etc.)</td>\n",
 392 |        "      <td>13</td>\n",
 393 |        "      <td>Employed full-time</td>\n",
 394 |        "      <td>116000.0</td>\n",
 395 |        "    </tr>\n",
 396 |        "    <tr>\n",
 397 |        "      <th>9</th>\n",
 398 |        "      <td>United Kingdom</td>\n",
 399 |        "      <td>Master’s degree (M.A., M.S., M.Eng., MBA, etc.)</td>\n",
 400 |        "      <td>4</td>\n",
 401 |        "      <td>Employed full-time</td>\n",
 402 |        "      <td>32315.0</td>\n",
 403 |        "    </tr>\n",
 404 |        "    <tr>\n",
 405 |        "      <th>10</th>\n",
 406 |        "      <td>United Kingdom</td>\n",
 407 |        "      <td>Bachelor’s degree (B.A., B.S., B.Eng., etc.)</td>\n",
 408 |        "      <td>2</td>\n",
 409 |        "      <td>Employed full-time</td>\n",
 410 |        "      <td>40070.0</td>\n",
 411 |        "    </tr>\n",
 412 |        "    <tr>\n",
 413 |        "      <th>11</th>\n",
 414 |        "      <td>Spain</td>\n",
 415 |        "      <td>Some college/university study without earning ...</td>\n",
 416 |        "      <td>7</td>\n",
 417 |        "      <td>Employed full-time</td>\n",
 418 |        "      <td>14268.0</td>\n",
 419 |        "    </tr>\n",
 420 |        "    <tr>\n",
 421 |        "      <th>12</th>\n",
 422 |        "      <td>Netherlands</td>\n",
 423 |        "      <td>Secondary school (e.g. American high school, G...</td>\n",
 424 |        "      <td>20</td>\n",
 425 |        "      <td>Employed full-time</td>\n",
 426 |        "      <td>38916.0</td>\n",
 427 |        "    </tr>\n",
 428 |        "  </tbody>\n",
 429 |        "</table>\n",
 430 |        "</div>"
 431 |       ],
 432 |       "text/plain": [
 433 |        "           Country                                            EdLevel  \\\n",
 434 |        "7    United States       Bachelor’s degree (B.A., B.S., B.Eng., etc.)   \n",
 435 |        "9   United Kingdom    Master’s degree (M.A., M.S., M.Eng., MBA, etc.)   \n",
 436 |        "10  United Kingdom       Bachelor’s degree (B.A., B.S., B.Eng., etc.)   \n",
 437 |        "11           Spain  Some college/university study without earning ...   \n",
 438 |        "12     Netherlands  Secondary school (e.g. American high school, G...   \n",
 439 |        "\n",
 440 |        "   YearsCodePro          Employment    Salary  \n",
 441 |        "7            13  Employed full-time  116000.0  \n",
 442 |        "9             4  Employed full-time   32315.0  \n",
 443 |        "10            2  Employed full-time   40070.0  \n",
 444 |        "11            7  Employed full-time   14268.0  \n",
 445 |        "12           20  Employed full-time   38916.0  "
 446 |       ]
 447 |      },
 448 |      "execution_count": 58,
 449 |      "metadata": {},
 450 |      "output_type": "execute_result"
 451 |     }
 452 |    ],
 453 |    "source": [
 454 |     "df = df[df[\"Salary\"].notnull()]\n",
 455 |     "df.head()"
 456 |    ]
 457 |   },
 458 |   {
 459 |    "cell_type": "code",
 460 |    "execution_count": 59,
 461 |    "id": "1c0a8af1",
 462 |    "metadata": {},
 463 |    "outputs": [
 464 |     {
 465 |      "name": "stdout",
 466 |      "output_type": "stream",
 467 |      "text": [
 468 |       "<class 'pandas.core.frame.DataFrame'>\n",
 469 |       "Int64Index: 34756 entries, 7 to 64154\n",
 470 |       "Data columns (total 5 columns):\n",
 471 |       " #   Column        Non-Null Count  Dtype  \n",
 472 |       "---  ------        --------------  -----  \n",
 473 |       " 0   Country       34756 non-null  object \n",
 474 |       " 1   EdLevel       34188 non-null  object \n",
 475 |       " 2   YearsCodePro  34621 non-null  object \n",
 476 |       " 3   Employment    34717 non-null  object \n",
 477 |       " 4   Salary        34756 non-null  float64\n",
 478 |       "dtypes: float64(1), object(4)\n",
 479 |       "memory usage: 1.6+ MB\n"
 480 |      ]
 481 |     }
 482 |    ],
 483 |    "source": [
 484 |     "df.info()"
 485 |    ]
 486 |   },
 487 |   {
 488 |    "cell_type": "code",
 489 |    "execution_count": 60,
 490 |    "id": "c1c2b6f7",
 491 |    "metadata": {},
 492 |    "outputs": [
 493 |     {
 494 |      "data": {
 495 |       "text/plain": [
 496 |        "Country         0\n",
 497 |        "EdLevel         0\n",
 498 |        "YearsCodePro    0\n",
 499 |        "Employment      0\n",
 500 |        "Salary          0\n",
 501 |        "dtype: int64"
 502 |       ]
 503 |      },
 504 |      "execution_count": 60,
 505 |      "metadata": {},
 506 |      "output_type": "execute_result"
 507 |     }
 508 |    ],
 509 |    "source": [
 510 |     "df = df.dropna()\n",
 511 |     "df.isnull().sum()"
 512 |    ]
 513 |   },
 514 |   {
 515 |    "cell_type": "code",
 516 |    "execution_count": 61,
 517 |    "id": "b876948a",
 518 |    "metadata": {},
 519 |    "outputs": [
 520 |     {
 521 |      "name": "stdout",
 522 |      "output_type": "stream",
 523 |      "text": [
 524 |       "<class 'pandas.core.frame.DataFrame'>\n",
 525 |       "Int64Index: 30019 entries, 7 to 64154\n",
 526 |       "Data columns (total 4 columns):\n",
 527 |       " #   Column        Non-Null Count  Dtype  \n",
 528 |       "---  ------        --------------  -----  \n",
 529 |       " 0   Country       30019 non-null  object \n",
 530 |       " 1   EdLevel       30019 non-null  object \n",
 531 |       " 2   YearsCodePro  30019 non-null  object \n",
 532 |       " 3   Salary        30019 non-null  float64\n",
 533 |       "dtypes: float64(1), object(3)\n",
 534 |       "memory usage: 1.1+ MB\n"
 535 |      ]
 536 |     }
 537 |    ],
 538 |    "source": [
 539 |     "df = df[df[\"Employment\"] == \"Employed full-time\"]\n",
 540 |     "df = df.drop(\"Employment\", axis=1)\n",
 541 |     "df.info()"
 542 |    ]
 543 |   },
 544 |   {
 545 |    "cell_type": "code",
 546 |    "execution_count": 62,
 547 |    "id": "c551ab05",
 548 |    "metadata": {},
 549 |    "outputs": [
 550 |     {
 551 |      "data": {
 552 |       "text/plain": [
 553 |        "United States                       7569\n",
 554 |        "India                               2425\n",
 555 |        "United Kingdom                      2287\n",
 556 |        "Germany                             1903\n",
 557 |        "Canada                              1178\n",
 558 |        "                                    ... \n",
 559 |        "San Marino                             1\n",
 560 |        "Saint Vincent and the Grenadines       1\n",
 561 |        "Namibia                                1\n",
 562 |        "Benin                                  1\n",
 563 |        "Monaco                                 1\n",
 564 |        "Name: Country, Length: 154, dtype: int64"
 565 |       ]
 566 |      },
 567 |      "execution_count": 62,
 568 |      "metadata": {},
 569 |      "output_type": "execute_result"
 570 |     }
 571 |    ],
 572 |    "source": [
 573 |     "df['Country'].value_counts()"
 574 |    ]
 575 |   },
 576 |   {
 577 |    "cell_type": "code",
 578 |    "execution_count": 63,
 579 |    "id": "cb8e3d2e",
 580 |    "metadata": {},
 581 |    "outputs": [],
 582 |    "source": [
 583 |     "def shorten_categories(categories, cutoff):\n",
 584 |     "    categorical_map = {}\n",
 585 |     "    for i in range(len(categories)):\n",
 586 |     "        if categories.values[i] >= cutoff:\n",
 587 |     "            categorical_map[categories.index[i]] = categories.index[i]\n",
 588 |     "        else:\n",
 589 |     "            categorical_map[categories.index[i]] = 'Other'\n",
 590 |     "    return categorical_map"
 591 |    ]
 592 |   },
 593 |   {
 594 |    "cell_type": "code",
 595 |    "execution_count": 64,
 596 |    "id": "cc09a5fe",
 597 |    "metadata": {},
 598 |    "outputs": [
 599 |     {
 600 |      "data": {
 601 |       "text/plain": [
 602 |        "Other                 8549\n",
 603 |        "United States         7569\n",
 604 |        "India                 2425\n",
 605 |        "United Kingdom        2287\n",
 606 |        "Germany               1903\n",
 607 |        "Canada                1178\n",
 608 |        "Brazil                 991\n",
 609 |        "France                 972\n",
 610 |        "Spain                  670\n",
 611 |        "Australia              659\n",
 612 |        "Netherlands            654\n",
 613 |        "Poland                 566\n",
 614 |        "Italy                  560\n",
 615 |        "Russian Federation     522\n",
 616 |        "Sweden                 514\n",
 617 |        "Name: Country, dtype: int64"
 618 |       ]
 619 |      },
 620 |      "execution_count": 64,
 621 |      "metadata": {},
 622 |      "output_type": "execute_result"
 623 |     }
 624 |    ],
 625 |    "source": [
 626 |     "country_map = shorten_categories(df.Country.value_counts(), 400)\n",
 627 |     "df['Country'] = df['Country'].map(country_map)\n",
 628 |     "df.Country.value_counts()"
 629 |    ]
 630 |   },
 631 |   {
 632 |    "cell_type": "code",
 633 |    "execution_count": 65,
 634 |    "id": "03a7116f",
 635 |    "metadata": {},
 636 |    "outputs": [
 637 |     {
 638 |      "data": {
 639 |       "image/png": "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\n",
 640 |       "text/plain": [
 641 |        "<Figure size 864x504 with 1 Axes>"
 642 |       ]
 643 |      },
 644 |      "metadata": {
 645 |       "needs_background": "light"
 646 |      },
 647 |      "output_type": "display_data"
 648 |     }
 649 |    ],
 650 |    "source": [
 651 |     "fig, ax = plt.subplots(1,1, figsize=(12, 7))\n",
 652 |     "df.boxplot('Salary', 'Country', ax=ax)\n",
 653 |     "plt.suptitle('Salary (US$) v Country')\n",
 654 |     "plt.title('')\n",
 655 |     "plt.ylabel('Salary')\n",
 656 |     "plt.xticks(rotation=90)\n",
 657 |     "plt.show()"
 658 |    ]
 659 |   },
 660 |   {
 661 |    "cell_type": "code",
 662 |    "execution_count": 66,
 663 |    "id": "29b6b862",
 664 |    "metadata": {},
 665 |    "outputs": [],
 666 |    "source": [
 667 |     "df = df[df[\"Salary\"] <= 250000]\n",
 668 |     "df = df[df[\"Salary\"] >= 10000]\n",
 669 |     "df = df[df['Country'] != 'Other']"
 670 |    ]
 671 |   },
 672 |   {
 673 |    "cell_type": "code",
 674 |    "execution_count": 67,
 675 |    "id": "ca55b532",
 676 |    "metadata": {},
 677 |    "outputs": [
 678 |     {
 679 |      "data": {
 680 |       "image/png": "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\n",
 681 |       "text/plain": [
 682 |        "<Figure size 864x504 with 1 Axes>"
 683 |       ]
 684 |      },
 685 |      "metadata": {
 686 |       "needs_background": "light"
 687 |      },
 688 |      "output_type": "display_data"
 689 |     }
 690 |    ],
 691 |    "source": [
 692 |     "fig, ax = plt.subplots(1,1, figsize=(12, 7))\n",
 693 |     "df.boxplot('Salary', 'Country', ax=ax)\n",
 694 |     "plt.suptitle('Salary (US$) v Country')\n",
 695 |     "plt.title('')\n",
 696 |     "plt.ylabel('Salary')\n",
 697 |     "plt.xticks(rotation=90)\n",
 698 |     "plt.show()"
 699 |    ]
 700 |   },
 701 |   {
 702 |    "cell_type": "code",
 703 |    "execution_count": 68,
 704 |    "id": "82cf4507",
 705 |    "metadata": {},
 706 |    "outputs": [
 707 |     {
 708 |      "data": {
 709 |       "text/plain": [
 710 |        "array(['13', '4', '2', '7', '20', '1', '3', '10', '12', '29', '6', '28',\n",
 711 |        "       '8', '23', '15', '25', '9', '11', 'Less than 1 year', '5', '21',\n",
 712 |        "       '16', '18', '14', '32', '19', '22', '38', '30', '26', '27', '17',\n",
 713 |        "       '24', '34', '35', '33', '36', '40', '39', 'More than 50 years',\n",
 714 |        "       '31', '37', '41', '45', '42', '44', '43', '50', '49'], dtype=object)"
 715 |       ]
 716 |      },
 717 |      "execution_count": 68,
 718 |      "metadata": {},
 719 |      "output_type": "execute_result"
 720 |     }
 721 |    ],
 722 |    "source": [
 723 |     "df[\"YearsCodePro\"].unique()"
 724 |    ]
 725 |   },
 726 |   {
 727 |    "cell_type": "code",
 728 |    "execution_count": 69,
 729 |    "id": "4f035a52",
 730 |    "metadata": {},
 731 |    "outputs": [],
 732 |    "source": [
 733 |     "def clean_experience(x):\n",
 734 |     "    if x ==  'More than 50 years':\n",
 735 |     "        return 50\n",
 736 |     "    if x == 'Less than 1 year':\n",
 737 |     "        return 0.5\n",
 738 |     "    return float(x)\n",
 739 |     "\n",
 740 |     "df['YearsCodePro'] = df['YearsCodePro'].apply(clean_experience)"
 741 |    ]
 742 |   },
 743 |   {
 744 |    "cell_type": "code",
 745 |    "execution_count": 70,
 746 |    "id": "c40b1ed5",
 747 |    "metadata": {},
 748 |    "outputs": [
 749 |     {
 750 |      "data": {
 751 |       "text/plain": [
 752 |        "array(['Bachelor’s degree (B.A., B.S., B.Eng., etc.)',\n",
 753 |        "       'Master’s degree (M.A., M.S., M.Eng., MBA, etc.)',\n",
 754 |        "       'Some college/university study without earning a degree',\n",
 755 |        "       'Secondary school (e.g. American high school, German Realschule or Gymnasium, etc.)',\n",
 756 |        "       'Associate degree (A.A., A.S., etc.)',\n",
 757 |        "       'Professional degree (JD, MD, etc.)',\n",
 758 |        "       'Other doctoral degree (Ph.D., Ed.D., etc.)',\n",
 759 |        "       'I never completed any formal education',\n",
 760 |        "       'Primary/elementary school'], dtype=object)"
 761 |       ]
 762 |      },
 763 |      "execution_count": 70,
 764 |      "metadata": {},
 765 |      "output_type": "execute_result"
 766 |     }
 767 |    ],
 768 |    "source": [
 769 |     "df[\"EdLevel\"].unique()"
 770 |    ]
 771 |   },
 772 |   {
 773 |    "cell_type": "code",
 774 |    "execution_count": 71,
 775 |    "id": "a24a1436",
 776 |    "metadata": {},
 777 |    "outputs": [],
 778 |    "source": [
 779 |     "def clean_education(x):\n",
 780 |     "    if 'Bachelor’s degree' in x:\n",
 781 |     "        return 'Bachelor’s degree'\n",
 782 |     "    if 'Master’s degree' in x:\n",
 783 |     "        return 'Master’s degree'\n",
 784 |     "    if 'Professional degree' in x or 'Other doctoral' in x:\n",
 785 |     "        return 'Post grad'\n",
 786 |     "    return 'Less than a Bachelors'\n",
 787 |     "\n",
 788 |     "df['EdLevel'] = df['EdLevel'].apply(clean_education)"
 789 |    ]
 790 |   },
 791 |   {
 792 |    "cell_type": "code",
 793 |    "execution_count": 72,
 794 |    "id": "9ce8792f",
 795 |    "metadata": {},
 796 |    "outputs": [
 797 |     {
 798 |      "data": {
 799 |       "text/plain": [
 800 |        "array(['Bachelor’s degree', 'Master’s degree', 'Less than a Bachelors',\n",
 801 |        "       'Post grad'], dtype=object)"
 802 |       ]
 803 |      },
 804 |      "execution_count": 72,
 805 |      "metadata": {},
 806 |      "output_type": "execute_result"
 807 |     }
 808 |    ],
 809 |    "source": [
 810 |     "df[\"EdLevel\"].unique()"
 811 |    ]
 812 |   },
 813 |   {
 814 |    "cell_type": "code",
 815 |    "execution_count": 73,
 816 |    "id": "896e4b84",
 817 |    "metadata": {},
 818 |    "outputs": [
 819 |     {
 820 |      "data": {
 821 |       "text/plain": [
 822 |        "array([0, 2, 1, 3])"
 823 |       ]
 824 |      },
 825 |      "execution_count": 73,
 826 |      "metadata": {},
 827 |      "output_type": "execute_result"
 828 |     }
 829 |    ],
 830 |    "source": [
 831 |     "from sklearn.preprocessing import LabelEncoder\n",
 832 |     "le_education = LabelEncoder()\n",
 833 |     "df['EdLevel'] = le_education.fit_transform(df['EdLevel'])\n",
 834 |     "df[\"EdLevel\"].unique()\n",
 835 |     "#le.classes_"
 836 |    ]
 837 |   },
 838 |   {
 839 |    "cell_type": "code",
 840 |    "execution_count": 74,
 841 |    "id": "3ee6c3b5",
 842 |    "metadata": {},
 843 |    "outputs": [
 844 |     {
 845 |      "data": {
 846 |       "text/plain": [
 847 |        "array([13, 12, 10,  7,  4,  2,  6,  1,  3,  5, 11,  8,  0,  9])"
 848 |       ]
 849 |      },
 850 |      "execution_count": 74,
 851 |      "metadata": {},
 852 |      "output_type": "execute_result"
 853 |     }
 854 |    ],
 855 |    "source": [
 856 |     "le_country = LabelEncoder()\n",
 857 |     "df['Country'] = le_country.fit_transform(df['Country'])\n",
 858 |     "df[\"Country\"].unique()"
 859 |    ]
 860 |   },
 861 |   {
 862 |    "cell_type": "code",
 863 |    "execution_count": 75,
 864 |    "id": "69a25de0",
 865 |    "metadata": {},
 866 |    "outputs": [],
 867 |    "source": [
 868 |     "X = df.drop(\"Salary\", axis=1)\n",
 869 |     "y = df[\"Salary\"]"
 870 |    ]
 871 |   },
 872 |   {
 873 |    "cell_type": "code",
 874 |    "execution_count": 39,
 875 |    "id": "8e1e3581",
 876 |    "metadata": {},
 877 |    "outputs": [
 878 |     {
 879 |      "data": {
 880 |       "text/plain": [
 881 |        "LinearRegression()"
 882 |       ]
 883 |      },
 884 |      "execution_count": 39,
 885 |      "metadata": {},
 886 |      "output_type": "execute_result"
 887 |     }
 888 |    ],
 889 |    "source": [
 890 |     "from sklearn.linear_model import LinearRegression\n",
 891 |     "linear_reg = LinearRegression()\n",
 892 |     "linear_reg.fit(X, y.values)"
 893 |    ]
 894 |   },
 895 |   {
 896 |    "cell_type": "code",
 897 |    "execution_count": 40,
 898 |    "id": "b8e531e2",
 899 |    "metadata": {},
 900 |    "outputs": [],
 901 |    "source": [
 902 |     "y_pred = linear_reg.predict(X)"
 903 |    ]
 904 |   },
 905 |   {
 906 |    "cell_type": "code",
 907 |    "execution_count": 41,
 908 |    "id": "5eb0fd40",
 909 |    "metadata": {},
 910 |    "outputs": [],
 911 |    "source": [
 912 |     "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
 913 |     "import numpy as np\n",
 914 |     "error = np.sqrt(mean_squared_error(y, y_pred))"
 915 |    ]
 916 |   },
 917 |   {
 918 |    "cell_type": "code",
 919 |    "execution_count": 42,
 920 |    "id": "8de6b173",
 921 |    "metadata": {},
 922 |    "outputs": [
 923 |     {
 924 |      "data": {
 925 |       "text/plain": [
 926 |        "39274.75368318509"
 927 |       ]
 928 |      },
 929 |      "execution_count": 42,
 930 |      "metadata": {},
 931 |      "output_type": "execute_result"
 932 |     }
 933 |    ],
 934 |    "source": [
 935 |     "error"
 936 |    ]
 937 |   },
 938 |   {
 939 |    "cell_type": "code",
 940 |    "execution_count": 43,
 941 |    "id": "2e633868",
 942 |    "metadata": {},
 943 |    "outputs": [
 944 |     {
 945 |      "data": {
 946 |       "text/plain": [
 947 |        "DecisionTreeRegressor(random_state=0)"
 948 |       ]
 949 |      },
 950 |      "execution_count": 43,
 951 |      "metadata": {},
 952 |      "output_type": "execute_result"
 953 |     }
 954 |    ],
 955 |    "source": [
 956 |     "from sklearn.tree import DecisionTreeRegressor\n",
 957 |     "dec_tree_reg = DecisionTreeRegressor(random_state=0)\n",
 958 |     "dec_tree_reg.fit(X, y.values)"
 959 |    ]
 960 |   },
 961 |   {
 962 |    "cell_type": "code",
 963 |    "execution_count": 44,
 964 |    "id": "9953f8bc",
 965 |    "metadata": {},
 966 |    "outputs": [],
 967 |    "source": [
 968 |     "y_pred = dec_tree_reg.predict(X)"
 969 |    ]
 970 |   },
 971 |   {
 972 |    "cell_type": "code",
 973 |    "execution_count": 45,
 974 |    "id": "806839af",
 975 |    "metadata": {},
 976 |    "outputs": [
 977 |     {
 978 |      "name": "stdout",
 979 |      "output_type": "stream",
 980 |      "text": [
 981 |       "$29,414.94\n"
 982 |      ]
 983 |     }
 984 |    ],
 985 |    "source": [
 986 |     "error = np.sqrt(mean_squared_error(y, y_pred))\n",
 987 |     "print(\"${:,.02f}\".format(error))"
 988 |    ]
 989 |   },
 990 |   {
 991 |    "cell_type": "code",
 992 |    "execution_count": 46,
 993 |    "id": "62d12912",
 994 |    "metadata": {},
 995 |    "outputs": [
 996 |     {
 997 |      "data": {
 998 |       "text/plain": [
 999 |        "RandomForestRegressor(random_state=0)"
1000 |       ]
1001 |      },
1002 |      "execution_count": 46,
1003 |      "metadata": {},
1004 |      "output_type": "execute_result"
1005 |     }
1006 |    ],
1007 |    "source": [
1008 |     "from sklearn.ensemble import RandomForestRegressor\n",
1009 |     "random_forest_reg = RandomForestRegressor(random_state=0)\n",
1010 |     "random_forest_reg.fit(X, y.values)"
1011 |    ]
1012 |   },
1013 |   {
1014 |    "cell_type": "code",
1015 |    "execution_count": 47,
1016 |    "id": "9a58e86e",
1017 |    "metadata": {},
1018 |    "outputs": [],
1019 |    "source": [
1020 |     "y_pred = random_forest_reg.predict(X)"
1021 |    ]
1022 |   },
1023 |   {
1024 |    "cell_type": "code",
1025 |    "execution_count": 48,
1026 |    "id": "b9c9f8a4",
1027 |    "metadata": {},
1028 |    "outputs": [
1029 |     {
1030 |      "name": "stdout",
1031 |      "output_type": "stream",
1032 |      "text": [
1033 |       "$29,487.31\n"
1034 |      ]
1035 |     }
1036 |    ],
1037 |    "source": [
1038 |     "error = np.sqrt(mean_squared_error(y, y_pred))\n",
1039 |     "print(\"${:,.02f}\".format(error))"
1040 |    ]
1041 |   },
1042 |   {
1043 |    "cell_type": "code",
1044 |    "execution_count": 49,
1045 |    "id": "36db7870",
1046 |    "metadata": {},
1047 |    "outputs": [
1048 |     {
1049 |      "data": {
1050 |       "text/plain": [
1051 |        "GridSearchCV(estimator=DecisionTreeRegressor(random_state=0),\n",
1052 |        "             param_grid={'max_depth': [None, 2, 4, 6, 8, 10, 12]},\n",
1053 |        "             scoring='neg_mean_squared_error')"
1054 |       ]
1055 |      },
1056 |      "execution_count": 49,
1057 |      "metadata": {},
1058 |      "output_type": "execute_result"
1059 |     }
1060 |    ],
1061 |    "source": [
1062 |     "from sklearn.model_selection import GridSearchCV\n",
1063 |     "\n",
1064 |     "max_depth = [None, 2,4,6,8,10,12]\n",
1065 |     "parameters = {\"max_depth\": max_depth}\n",
1066 |     "\n",
1067 |     "regressor = DecisionTreeRegressor(random_state=0)\n",
1068 |     "gs = GridSearchCV(regressor, parameters, scoring='neg_mean_squared_error')\n",
1069 |     "gs.fit(X, y.values)"
1070 |    ]
1071 |   },
1072 |   {
1073 |    "cell_type": "code",
1074 |    "execution_count": 50,
1075 |    "id": "11fddae1",
1076 |    "metadata": {},
1077 |    "outputs": [
1078 |     {
1079 |      "name": "stdout",
1080 |      "output_type": "stream",
1081 |      "text": [
1082 |       "$30,428.51\n"
1083 |      ]
1084 |     }
1085 |    ],
1086 |    "source": [
1087 |     "regressor = gs.best_estimator_\n",
1088 |     "\n",
1089 |     "regressor.fit(X, y.values)\n",
1090 |     "y_pred = regressor.predict(X)\n",
1091 |     "error = np.sqrt(mean_squared_error(y, y_pred))\n",
1092 |     "print(\"${:,.02f}\".format(error))"
1093 |    ]
1094 |   },
1095 |   {
1096 |    "cell_type": "code",
1097 |    "execution_count": 76,
1098 |    "id": "d1c7b5ac",
1099 |    "metadata": {},
1100 |    "outputs": [
1101 |     {
1102 |      "data": {
1103 |       "text/html": [
1104 |        "<div>\n",
1105 |        "<style scoped>\n",
1106 |        "    .dataframe tbody tr th:only-of-type {\n",
1107 |        "        vertical-align: middle;\n",
1108 |        "    }\n",
1109 |        "\n",
1110 |        "    .dataframe tbody tr th {\n",
1111 |        "        vertical-align: top;\n",
1112 |        "    }\n",
1113 |        "\n",
1114 |        "    .dataframe thead th {\n",
1115 |        "        text-align: right;\n",
1116 |        "    }\n",
1117 |        "</style>\n",
1118 |        "<table border=\"1\" class=\"dataframe\">\n",
1119 |        "  <thead>\n",
1120 |        "    <tr style=\"text-align: right;\">\n",
1121 |        "      <th></th>\n",
1122 |        "      <th>Country</th>\n",
1123 |        "      <th>EdLevel</th>\n",
1124 |        "      <th>YearsCodePro</th>\n",
1125 |        "    </tr>\n",
1126 |        "  </thead>\n",
1127 |        "  <tbody>\n",
1128 |        "    <tr>\n",
1129 |        "      <th>7</th>\n",
1130 |        "      <td>13</td>\n",
1131 |        "      <td>0</td>\n",
1132 |        "      <td>13.0</td>\n",
1133 |        "    </tr>\n",
1134 |        "    <tr>\n",
1135 |        "      <th>9</th>\n",
1136 |        "      <td>12</td>\n",
1137 |        "      <td>2</td>\n",
1138 |        "      <td>4.0</td>\n",
1139 |        "    </tr>\n",
1140 |        "    <tr>\n",
1141 |        "      <th>10</th>\n",
1142 |        "      <td>12</td>\n",
1143 |        "      <td>0</td>\n",
1144 |        "      <td>2.0</td>\n",
1145 |        "    </tr>\n",
1146 |        "    <tr>\n",
1147 |        "      <th>11</th>\n",
1148 |        "      <td>10</td>\n",
1149 |        "      <td>1</td>\n",
1150 |        "      <td>7.0</td>\n",
1151 |        "    </tr>\n",
1152 |        "    <tr>\n",
1153 |        "      <th>12</th>\n",
1154 |        "      <td>7</td>\n",
1155 |        "      <td>1</td>\n",
1156 |        "      <td>20.0</td>\n",
1157 |        "    </tr>\n",
1158 |        "    <tr>\n",
1159 |        "      <th>...</th>\n",
1160 |        "      <td>...</td>\n",
1161 |        "      <td>...</td>\n",
1162 |        "      <td>...</td>\n",
1163 |        "    </tr>\n",
1164 |        "    <tr>\n",
1165 |        "      <th>64113</th>\n",
1166 |        "      <td>13</td>\n",
1167 |        "      <td>1</td>\n",
1168 |        "      <td>15.0</td>\n",
1169 |        "    </tr>\n",
1170 |        "    <tr>\n",
1171 |        "      <th>64116</th>\n",
1172 |        "      <td>13</td>\n",
1173 |        "      <td>0</td>\n",
1174 |        "      <td>6.0</td>\n",
1175 |        "    </tr>\n",
1176 |        "    <tr>\n",
1177 |        "      <th>64122</th>\n",
1178 |        "      <td>13</td>\n",
1179 |        "      <td>1</td>\n",
1180 |        "      <td>4.0</td>\n",
1181 |        "    </tr>\n",
1182 |        "    <tr>\n",
1183 |        "      <th>64127</th>\n",
1184 |        "      <td>13</td>\n",
1185 |        "      <td>3</td>\n",
1186 |        "      <td>12.0</td>\n",
1187 |        "    </tr>\n",
1188 |        "    <tr>\n",
1189 |        "      <th>64129</th>\n",
1190 |        "      <td>13</td>\n",
1191 |        "      <td>2</td>\n",
1192 |        "      <td>4.0</td>\n",
1193 |        "    </tr>\n",
1194 |        "  </tbody>\n",
1195 |        "</table>\n",
1196 |        "<p>18491 rows × 3 columns</p>\n",
1197 |        "</div>"
1198 |       ],
1199 |       "text/plain": [
1200 |        "       Country  EdLevel  YearsCodePro\n",
1201 |        "7           13        0          13.0\n",
1202 |        "9           12        2           4.0\n",
1203 |        "10          12        0           2.0\n",
1204 |        "11          10        1           7.0\n",
1205 |        "12           7        1          20.0\n",
1206 |        "...        ...      ...           ...\n",
1207 |        "64113       13        1          15.0\n",
1208 |        "64116       13        0           6.0\n",
1209 |        "64122       13        1           4.0\n",
1210 |        "64127       13        3          12.0\n",
1211 |        "64129       13        2           4.0\n",
1212 |        "\n",
1213 |        "[18491 rows x 3 columns]"
1214 |       ]
1215 |      },
1216 |      "execution_count": 76,
1217 |      "metadata": {},
1218 |      "output_type": "execute_result"
1219 |     }
1220 |    ],
1221 |    "source": [
1222 |     "X"
1223 |    ]
1224 |   },
1225 |   {
1226 |    "cell_type": "code",
1227 |    "execution_count": 77,
1228 |    "id": "c947101a",
1229 |    "metadata": {},
1230 |    "outputs": [
1231 |     {
1232 |      "data": {
1233 |       "text/plain": [
1234 |        "array([['United States', 'Master’s degree', '15']], dtype='<U21')"
1235 |       ]
1236 |      },
1237 |      "execution_count": 77,
1238 |      "metadata": {},
1239 |      "output_type": "execute_result"
1240 |     }
1241 |    ],
1242 |    "source": [
1243 |     "# country, edlevel, yearscode\n",
1244 |     "X = np.array([[\"United States\", 'Master’s degree', 15 ]])\n",
1245 |     "X"
1246 |    ]
1247 |   },
1248 |   {
1249 |    "cell_type": "code",
1250 |    "execution_count": 78,
1251 |    "id": "914062f3",
1252 |    "metadata": {},
1253 |    "outputs": [
1254 |     {
1255 |      "data": {
1256 |       "text/plain": [
1257 |        "array([[13.,  2., 15.]])"
1258 |       ]
1259 |      },
1260 |      "execution_count": 78,
1261 |      "metadata": {},
1262 |      "output_type": "execute_result"
1263 |     }
1264 |    ],
1265 |    "source": [
1266 |     "X[:, 0] = le_country.transform(X[:,0])\n",
1267 |     "X[:, 1] = le_education.transform(X[:,1])\n",
1268 |     "X = X.astype(float)\n",
1269 |     "X"
1270 |    ]
1271 |   },
1272 |   {
1273 |    "cell_type": "code",
1274 |    "execution_count": 79,
1275 |    "id": "258d1fab",
1276 |    "metadata": {},
1277 |    "outputs": [
1278 |     {
1279 |      "data": {
1280 |       "text/plain": [
1281 |        "array([139427.26315789])"
1282 |       ]
1283 |      },
1284 |      "execution_count": 79,
1285 |      "metadata": {},
1286 |      "output_type": "execute_result"
1287 |     }
1288 |    ],
1289 |    "source": [
1290 |     "y_pred = regressor.predict(X)\n",
1291 |     "y_pred"
1292 |    ]
1293 |   },
1294 |   {
1295 |    "cell_type": "code",
1296 |    "execution_count": 80,
1297 |    "id": "a1203e5c",
1298 |    "metadata": {},
1299 |    "outputs": [],
1300 |    "source": [
1301 |     "import pickle"
1302 |    ]
1303 |   },
1304 |   {
1305 |    "cell_type": "code",
1306 |    "execution_count": 81,
1307 |    "id": "6fc64774",
1308 |    "metadata": {},
1309 |    "outputs": [],
1310 |    "source": [
1311 |     "data = {\"model\": regressor, \"le_country\": le_country, \"le_education\": le_education}\n",
1312 |     "with open('saved_steps.pkl', 'wb') as file:\n",
1313 |     "    pickle.dump(data, file)"
1314 |    ]
1315 |   },
1316 |   {
1317 |    "cell_type": "code",
1318 |    "execution_count": 82,
1319 |    "id": "6aefe42c",
1320 |    "metadata": {},
1321 |    "outputs": [],
1322 |    "source": [
1323 |     "with open('saved_steps.pkl', 'rb') as file:\n",
1324 |     "    data = pickle.load(file)\n",
1325 |     "\n",
1326 |     "regressor_loaded = data[\"model\"]\n",
1327 |     "le_country = data[\"le_country\"]\n",
1328 |     "le_education = data[\"le_education\"]"
1329 |    ]
1330 |   },
1331 |   {
1332 |    "cell_type": "code",
1333 |    "execution_count": 83,
1334 |    "id": "eeb1983b",
1335 |    "metadata": {},
1336 |    "outputs": [
1337 |     {
1338 |      "data": {
1339 |       "text/plain": [
1340 |        "array([139427.26315789])"
1341 |       ]
1342 |      },
1343 |      "execution_count": 83,
1344 |      "metadata": {},
1345 |      "output_type": "execute_result"
1346 |     }
1347 |    ],
1348 |    "source": [
1349 |     "y_pred = regressor_loaded.predict(X)\n",
1350 |     "y_pred"
1351 |    ]
1352 |   },
1353 |   {
1354 |    "cell_type": "code",
1355 |    "execution_count": null,
1356 |    "id": "375424e4",
1357 |    "metadata": {},
1358 |    "outputs": [],
1359 |    "source": [
1360 |     " "
1361 |    ]
1362 |   },
1363 |   {
1364 |    "cell_type": "code",
1365 |    "execution_count": null,
1366 |    "id": "1105963a",
1367 |    "metadata": {},
1368 |    "outputs": [],
1369 |    "source": []
1370 |   }
1371 |  ],
1372 |  "metadata": {
1373 |   "kernelspec": {
1374 |    "display_name": "ml",
1375 |    "language": "python",
1376 |    "name": "ml"
1377 |   },
1378 |   "language_info": {
1379 |    "codemirror_mode": {
1380 |     "name": "ipython",
1381 |     "version": 3
1382 |    },
1383 |    "file_extension": ".py",
1384 |    "mimetype": "text/x-python",
1385 |    "name": "python",
1386 |    "nbconvert_exporter": "python",
1387 |    "pygments_lexer": "ipython3",
1388 |    "version": "3.9.2"
1389 |   }
1390 |  },
1391 |  "nbformat": 4,
1392 |  "nbformat_minor": 5
1393 | }
1394 | 


--------------------------------------------------------------------------------
/app.py:
--------------------------------------------------------------------------------
 1 | import streamlit as st
 2 | from predict_page import show_predict_page
 3 | from explore_page import show_explore_page
 4 | 
 5 | 
 6 | page = st.sidebar.selectbox("Explore Or Predict", ("Predict", "Explore"))
 7 | 
 8 | if page == "Predict":
 9 |     show_predict_page()
10 | else:
11 |     show_explore_page()
12 | 


--------------------------------------------------------------------------------
/explore_page.py:
--------------------------------------------------------------------------------
 1 | import streamlit as st
 2 | import pandas as pd
 3 | import matplotlib.pyplot as plt
 4 | 
 5 | def shorten_categories(categories, cutoff):
 6 |     categorical_map = {}
 7 |     for i in range(len(categories)):
 8 |         if categories.values[i] >= cutoff:
 9 |             categorical_map[categories.index[i]] = categories.index[i]
10 |         else:
11 |             categorical_map[categories.index[i]] = 'Other'
12 |     return categorical_map
13 | 
14 | 
15 | def clean_experience(x):
16 |     if x ==  'More than 50 years':
17 |         return 50
18 |     if x == 'Less than 1 year':
19 |         return 0.5
20 |     return float(x)
21 | 
22 | 
23 | def clean_education(x):
24 |     if 'Bachelor’s degree' in x:
25 |         return 'Bachelor’s degree'
26 |     if 'Master’s degree' in x:
27 |         return 'Master’s degree'
28 |     if 'Professional degree' in x or 'Other doctoral' in x:
29 |         return 'Post grad'
30 |     return 'Less than a Bachelors'
31 | 
32 | 
33 | @st.cache
34 | def load_data():
35 |     df = pd.read_csv("survey_results_public.csv")
36 |     df = df[["Country", "EdLevel", "YearsCodePro", "Employment", "ConvertedComp"]]
37 |     df = df[df["ConvertedComp"].notnull()]
38 |     df = df.dropna()
39 |     df = df[df["Employment"] == "Employed full-time"]
40 |     df = df.drop("Employment", axis=1)
41 | 
42 |     country_map = shorten_categories(df.Country.value_counts(), 400)
43 |     df["Country"] = df["Country"].map(country_map)
44 |     df = df[df["ConvertedComp"] <= 250000]
45 |     df = df[df["ConvertedComp"] >= 10000]
46 |     df = df[df["Country"] != "Other"]
47 | 
48 |     df["YearsCodePro"] = df["YearsCodePro"].apply(clean_experience)
49 |     df["EdLevel"] = df["EdLevel"].apply(clean_education)
50 |     df = df.rename({"ConvertedComp": "Salary"}, axis=1)
51 |     return df
52 | 
53 | df = load_data()
54 | 
55 | def show_explore_page():
56 |     st.title("Explore Software Engineer Salaries")
57 | 
58 |     st.write(
59 |         """
60 |     ### Stack Overflow Developer Survey 2020
61 |     """
62 |     )
63 | 
64 |     data = df["Country"].value_counts()
65 | 
66 |     fig1, ax1 = plt.subplots()
67 |     ax1.pie(data, labels=data.index, autopct="%1.1f%%", shadow=True, startangle=90)
68 |     ax1.axis("equal")  # Equal aspect ratio ensures that pie is drawn as a circle.
69 | 
70 |     st.write("""#### Number of Data from different countries""")
71 | 
72 |     st.pyplot(fig1)
73 |     
74 |     st.write(
75 |         """
76 |     #### Mean Salary Based On Country
77 |     """
78 |     )
79 | 
80 |     data = df.groupby(["Country"])["Salary"].mean().sort_values(ascending=True)
81 |     st.bar_chart(data)
82 | 
83 |     st.write(
84 |         """
85 |     #### Mean Salary Based On Experience
86 |     """
87 |     )
88 | 
89 |     data = df.groupby(["YearsCodePro"])["Salary"].mean().sort_values(ascending=True)
90 |     st.line_chart(data)
91 | 
92 | 


--------------------------------------------------------------------------------
/predict_page.py:
--------------------------------------------------------------------------------
 1 | import streamlit as st
 2 | import pickle
 3 | import numpy as np
 4 | 
 5 | 
 6 | def load_model():
 7 |     with open('saved_steps.pkl', 'rb') as file:
 8 |         data = pickle.load(file)
 9 |     return data
10 | 
11 | data = load_model()
12 | 
13 | regressor = data["model"]
14 | le_country = data["le_country"]
15 | le_education = data["le_education"]
16 | 
17 | def show_predict_page():
18 |     st.title("Software Developer Salary Prediction")
19 | 
20 |     st.write("""### We need some information to predict the salary""")
21 | 
22 |     countries = (
23 |         "United States",
24 |         "India",
25 |         "United Kingdom",
26 |         "Germany",
27 |         "Canada",
28 |         "Brazil",
29 |         "France",
30 |         "Spain",
31 |         "Australia",
32 |         "Netherlands",
33 |         "Poland",
34 |         "Italy",
35 |         "Russian Federation",
36 |         "Sweden",
37 |     )
38 | 
39 |     education = (
40 |         "Less than a Bachelors",
41 |         "Bachelor’s degree",
42 |         "Master’s degree",
43 |         "Post grad",
44 |     )
45 | 
46 |     country = st.selectbox("Country", countries)
47 |     education = st.selectbox("Education Level", education)
48 | 
49 |     expericence = st.slider("Years of Experience", 0, 50, 3)
50 | 
51 |     ok = st.button("Calculate Salary")
52 |     if ok:
53 |         X = np.array([[country, education, expericence ]])
54 |         X[:, 0] = le_country.transform(X[:,0])
55 |         X[:, 1] = le_education.transform(X[:,1])
56 |         X = X.astype(float)
57 | 
58 |         salary = regressor.predict(X)
59 |         st.subheader(f"The estimated salary is ${salary[0]:.2f}")
60 | 


--------------------------------------------------------------------------------