├── .gitignore
├── README.md
├── data
    └── interviews.csv
└── just-pandas-things.ipynb


/.gitignore:
--------------------------------------------------------------------------------
1 | .ipynb_checkpoints
2 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # just-pandas-things
 2 | This repo contains a few peculiar things I've learned about pandas that have made my life easier and my code faster. This post isn't a friendly tutorial for beginners, but a friendly introduction to pandas weirdness.
 3 | 
 4 | ## What's in this repo?
 5 | 
 6 | 1. pandas is column-major, which is why row-based operations are slow
 7 | 2. `SettingWithCopyWarning`, or why we can't have nice things
 8 | 3. Indexing and slicing
 9 | 4. Accessors
10 | 5. Data exploration
11 | 6. Common pitfalls
12 | 
13 | I'll continue updating this repo as I have more time. As I'm still learning pandas quirks, feedback is much appreciated.
14 | 
15 | Thanks [Luke Metz](https://twitter.com/luke_metz), [Vikram Tiwari](https://twitter.com/Vikram_Tiwari), and [Karson Elmgren](https://karsonelmgren.com/) for reviewing!
16 | 


--------------------------------------------------------------------------------
/just-pandas-things.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |  "cells": [
   3 |   {
   4 |    "cell_type": "markdown",
   5 |    "metadata": {},
   6 |    "source": [
   7 |     "# Just pandas things\n",
   8 |     "\n",
   9 |     "It's possible that Python wouldn't have become [the lingua franca of data science if it wasn't for pandas](https://stackoverflow.blog/2017/09/14/python-growing-quickly/). The package's exponential growth on Stack Overflow means two things:\n",
  10 |     "1. It's getting increasingly popular.\n",
  11 |     "2. It can be frustrating to use sometimes (hence the high number of questions).\n",
  12 |     "\n",
  13 |     "This repo contains a few peculiar things I've learned about pandas that have made my life easier and my code faster. This post isn't a friendly tutorial for beginners, but a friendly introduction to pandas weirdness.\n",
  14 |     "\n",
  15 |     "To demonstrate the use of pandas, we'll be using the interview reviews scraped from Glassdoor in 2019. The data is stored in the folder `data` under the name `interviews.csv`.\n",
  16 |     "\n",
  17 |     "I'll continue updating this repo as I have more time. As I'm still learning pandas quirks, feedback is much appreciated!\n",
  18 |     "\n",
  19 |     "![](https://i.pinimg.com/originals/9e/7c/78/9e7c7816c30327890dc94ba16e5dac1b.jpg)"
  20 |    ]
  21 |   },
  22 |   {
  23 |    "cell_type": "code",
  24 |    "execution_count": 1,
  25 |    "metadata": {},
  26 |    "outputs": [
  27 |     {
  28 |      "name": "stdout",
  29 |      "output_type": "stream",
  30 |      "text": [
  31 |       "(17654, 10)\n"
  32 |      ]
  33 |     },
  34 |     {
  35 |      "data": {
  36 |       "text/html": [
  37 |        "<div>\n",
  38 |        "<style scoped>\n",
  39 |        "    .dataframe tbody tr th:only-of-type {\n",
  40 |        "        vertical-align: middle;\n",
  41 |        "    }\n",
  42 |        "\n",
  43 |        "    .dataframe tbody tr th {\n",
  44 |        "        vertical-align: top;\n",
  45 |        "    }\n",
  46 |        "\n",
  47 |        "    .dataframe thead th {\n",
  48 |        "        text-align: right;\n",
  49 |        "    }\n",
  50 |        "</style>\n",
  51 |        "<table border=\"1\" class=\"dataframe\">\n",
  52 |        "  <thead>\n",
  53 |        "    <tr style=\"text-align: right;\">\n",
  54 |        "      <th></th>\n",
  55 |        "      <th>Company</th>\n",
  56 |        "      <th>Title</th>\n",
  57 |        "      <th>Job</th>\n",
  58 |        "      <th>Level</th>\n",
  59 |        "      <th>Date</th>\n",
  60 |        "      <th>Upvotes</th>\n",
  61 |        "      <th>Offer</th>\n",
  62 |        "      <th>Experience</th>\n",
  63 |        "      <th>Difficulty</th>\n",
  64 |        "      <th>Review</th>\n",
  65 |        "    </tr>\n",
  66 |        "  </thead>\n",
  67 |        "  <tbody>\n",
  68 |        "    <tr>\n",
  69 |        "      <th>0</th>\n",
  70 |        "      <td>Apple</td>\n",
  71 |        "      <td>Software Engineer</td>\n",
  72 |        "      <td>Software Engineer</td>\n",
  73 |        "      <td>Engineer</td>\n",
  74 |        "      <td>Aug 7, 2019</td>\n",
  75 |        "      <td>0</td>\n",
  76 |        "      <td>No offer</td>\n",
  77 |        "      <td>0.0</td>\n",
  78 |        "      <td>Medium</td>\n",
  79 |        "      <td>Application  I applied through a staffing agen...</td>\n",
  80 |        "    </tr>\n",
  81 |        "    <tr>\n",
  82 |        "      <th>1</th>\n",
  83 |        "      <td>Apple</td>\n",
  84 |        "      <td>Software Engineer</td>\n",
  85 |        "      <td>Software Engineer</td>\n",
  86 |        "      <td>Engineer</td>\n",
  87 |        "      <td>Aug 8, 2019</td>\n",
  88 |        "      <td>0</td>\n",
  89 |        "      <td>Accepted offer</td>\n",
  90 |        "      <td>1.0</td>\n",
  91 |        "      <td>Hard</td>\n",
  92 |        "      <td>Application  I applied online. The process too...</td>\n",
  93 |        "    </tr>\n",
  94 |        "    <tr>\n",
  95 |        "      <th>2</th>\n",
  96 |        "      <td>Apple</td>\n",
  97 |        "      <td>Software Engineer</td>\n",
  98 |        "      <td>Software Engineer</td>\n",
  99 |        "      <td>Engineer</td>\n",
 100 |        "      <td>NaN</td>\n",
 101 |        "      <td>0</td>\n",
 102 |        "      <td>Declined offer</td>\n",
 103 |        "      <td>0.0</td>\n",
 104 |        "      <td>Medium</td>\n",
 105 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
 106 |        "    </tr>\n",
 107 |        "    <tr>\n",
 108 |        "      <th>3</th>\n",
 109 |        "      <td>Apple</td>\n",
 110 |        "      <td>Software Engineer</td>\n",
 111 |        "      <td>Software Engineer</td>\n",
 112 |        "      <td>Engineer</td>\n",
 113 |        "      <td>NaN</td>\n",
 114 |        "      <td>9</td>\n",
 115 |        "      <td>Declined offer</td>\n",
 116 |        "      <td>-1.0</td>\n",
 117 |        "      <td>Medium</td>\n",
 118 |        "      <td>Application  The process took a week. I interv...</td>\n",
 119 |        "    </tr>\n",
 120 |        "    <tr>\n",
 121 |        "      <th>4</th>\n",
 122 |        "      <td>Apple</td>\n",
 123 |        "      <td>Software Engineer</td>\n",
 124 |        "      <td>Software Engineer</td>\n",
 125 |        "      <td>Engineer</td>\n",
 126 |        "      <td>May 29, 2009</td>\n",
 127 |        "      <td>2</td>\n",
 128 |        "      <td>No offer</td>\n",
 129 |        "      <td>0.0</td>\n",
 130 |        "      <td>Medium</td>\n",
 131 |        "      <td>Application  I applied through an employee ref...</td>\n",
 132 |        "    </tr>\n",
 133 |        "  </tbody>\n",
 134 |        "</table>\n",
 135 |        "</div>"
 136 |       ],
 137 |       "text/plain": [
 138 |        "  Company              Title                Job     Level          Date  \\\n",
 139 |        "0   Apple  Software Engineer  Software Engineer  Engineer   Aug 7, 2019   \n",
 140 |        "1   Apple  Software Engineer  Software Engineer  Engineer   Aug 8, 2019   \n",
 141 |        "2   Apple  Software Engineer  Software Engineer  Engineer           NaN   \n",
 142 |        "3   Apple  Software Engineer  Software Engineer  Engineer           NaN   \n",
 143 |        "4   Apple  Software Engineer  Software Engineer  Engineer  May 29, 2009   \n",
 144 |        "\n",
 145 |        "   Upvotes           Offer  Experience Difficulty  \\\n",
 146 |        "0        0        No offer         0.0     Medium   \n",
 147 |        "1        0  Accepted offer         1.0       Hard   \n",
 148 |        "2        0  Declined offer         0.0     Medium   \n",
 149 |        "3        9  Declined offer        -1.0     Medium   \n",
 150 |        "4        2        No offer         0.0     Medium   \n",
 151 |        "\n",
 152 |        "                                              Review  \n",
 153 |        "0  Application  I applied through a staffing agen...  \n",
 154 |        "1  Application  I applied online. The process too...  \n",
 155 |        "2  Application  The process took 4 weeks. I inter...  \n",
 156 |        "3  Application  The process took a week. I interv...  \n",
 157 |        "4  Application  I applied through an employee ref...  "
 158 |       ]
 159 |      },
 160 |      "execution_count": 1,
 161 |      "metadata": {},
 162 |      "output_type": "execute_result"
 163 |     }
 164 |    ],
 165 |    "source": [
 166 |     "import pandas as pd\n",
 167 |     "df = pd.read_csv(\"data/interviews.csv\")\n",
 168 |     "\n",
 169 |     "print(df.shape)\n",
 170 |     "df.head()"
 171 |    ]
 172 |   },
 173 |   {
 174 |    "cell_type": "markdown",
 175 |    "metadata": {},
 176 |    "source": [
 177 |     "## 1. pandas is column-major\n",
 178 |     "Pandas is built around `DataFrame`, a concept inspired by R's Data Frame, which is, in turn, similar to tables in relational databases. A `DataFrame` is a two-dimensional table with rows and columns.\n",
 179 |     "\n",
 180 |     "One important thing to know about pandas is that it's column-major, which explains many of its quirks.\n",
 181 |     "\n",
 182 |     "Column-major means consecutive elements in a column are stored next to each other in memory. Row-major means the same but for elements in a row. Because modern computers process sequential data more efficiently than nonsequential data, if a table is row-major, accessing its rows will be much faster than accessing its columns.\n",
 183 |     "\n",
 184 |     "In NumPy, major order can be specified. When a `ndarray` is created, it's row-major by default if you don't specify the order.\n",
 185 |     "\n",
 186 |     "Like R's Data Frame, pandas' `DataFrame` is column-major. People coming to pandas from NumPy tend to treat `DataFrame` the way they would `ndarray`, e.g. trying to access data by rows, and find `DataFrame` slow.\n",
 187 |     "\n",
 188 |     "**Note**: A column in a `DataFrame` is a `Series`. You can think of a `DataFrame` as a bunch of `Series` being stored next to each other in memory.\n",
 189 |     "\n",
 190 |     "**For our dataset, accessing a row takes about 50x longer than accessing a column in our `DataFrame`.**"
 191 |    ]
 192 |   },
 193 |   {
 194 |    "cell_type": "code",
 195 |    "execution_count": 2,
 196 |    "metadata": {},
 197 |    "outputs": [
 198 |     {
 199 |      "name": "stdout",
 200 |      "output_type": "stream",
 201 |      "text": [
 202 |       "1.78 µs ± 167 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
 203 |       "145 µs ± 9.41 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n"
 204 |      ]
 205 |     }
 206 |    ],
 207 |    "source": [
 208 |     "# Get the column `date`, 1000 loops\n",
 209 |     "%timeit -n1000 df[\"Date\"]\n",
 210 |     "\n",
 211 |     "# Get the first row, 1000 loops\n",
 212 |     "%timeit -n1000 df.iloc[0]"
 213 |    ]
 214 |   },
 215 |   {
 216 |    "cell_type": "markdown",
 217 |    "metadata": {},
 218 |    "source": [
 219 |     "### 1.1 Iterating over rows\n",
 220 |     "#### 1.1.1 `.apply()`\n",
 221 |     "pandas documentation has [a warning box](https://pandas.pydata.org/pandas-docs/stable/user_guide/basics.html#iteration) that basically tells you not to iterate over rows because it's slow.\n",
 222 |     "\n",
 223 |     "Before iterating over rows, think about what you want to do with each row, pack that into a function and use methods like `.apply()` to apply the function to all rows.\n",
 224 |     "\n",
 225 |     "For example, to scale the \"Experience\" column by the number of \"Upvotes\" each review has, one way is to iterate over rows and multiple the \"Upvotes\" value by the \"Experience\" value of that row. But you can also use `.apply()` with a `lambda` function."
 226 |    ]
 227 |   },
 228 |   {
 229 |    "cell_type": "code",
 230 |    "execution_count": 3,
 231 |    "metadata": {},
 232 |    "outputs": [
 233 |     {
 234 |      "name": "stdout",
 235 |      "output_type": "stream",
 236 |      "text": [
 237 |       "180 ms ± 16.5 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
 238 |      ]
 239 |     }
 240 |    ],
 241 |    "source": [
 242 |     "%timeit -n1 df.apply(lambda x: x[\"Experience\"] * x[\"Upvotes\"], axis=1)"
 243 |    ]
 244 |   },
 245 |   {
 246 |    "cell_type": "markdown",
 247 |    "metadata": {},
 248 |    "source": [
 249 |     "#### 1.1.2 `.iterrows()` and `.itertuples()`\n",
 250 |     "If you really want to iterate over rows, one naive way is to use `.iterrows()`. It returns a generator that generates row by row and it's very slow."
 251 |    ]
 252 |   },
 253 |   {
 254 |    "cell_type": "code",
 255 |    "execution_count": 4,
 256 |    "metadata": {},
 257 |    "outputs": [
 258 |     {
 259 |      "name": "stdout",
 260 |      "output_type": "stream",
 261 |      "text": [
 262 |       "1.42 s ± 107 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
 263 |      ]
 264 |     }
 265 |    ],
 266 |    "source": [
 267 |     "%timeit -n1 [row for index, row in df.iterrows()]"
 268 |    ]
 269 |   },
 270 |   {
 271 |    "cell_type": "code",
 272 |    "execution_count": 5,
 273 |    "metadata": {},
 274 |    "outputs": [
 275 |     {
 276 |      "name": "stdout",
 277 |      "output_type": "stream",
 278 |      "text": [
 279 |       "Company                                                   Apple\n",
 280 |       "Title                                         Software Engineer\n",
 281 |       "Job                                           Software Engineer\n",
 282 |       "Level                                                  Engineer\n",
 283 |       "Date                                                Aug 7, 2019\n",
 284 |       "Upvotes                                                       0\n",
 285 |       "Offer                                                  No offer\n",
 286 |       "Experience                                                    0\n",
 287 |       "Difficulty                                               Medium\n",
 288 |       "Review        Application  I applied through a staffing agen...\n",
 289 |       "Name: 0, dtype: object\n"
 290 |      ]
 291 |     }
 292 |    ],
 293 |    "source": [
 294 |     "# This is what a row looks like as a pandas object\n",
 295 |     "for index, row in df.iterrows():\n",
 296 |     "    print(row)\n",
 297 |     "    break"
 298 |    ]
 299 |   },
 300 |   {
 301 |    "cell_type": "markdown",
 302 |    "metadata": {},
 303 |    "source": [
 304 |     "`.itertuples()` returns rows in the `namedtuple` format. It still lets you access each row and it's about 40x faster than `.iterrows()`."
 305 |    ]
 306 |   },
 307 |   {
 308 |    "cell_type": "code",
 309 |    "execution_count": 6,
 310 |    "metadata": {},
 311 |    "outputs": [
 312 |     {
 313 |      "name": "stdout",
 314 |      "output_type": "stream",
 315 |      "text": [
 316 |       "24.2 ms ± 709 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
 317 |      ]
 318 |     }
 319 |    ],
 320 |    "source": [
 321 |     "%timeit -n1 [row for row in df.itertuples()]"
 322 |    ]
 323 |   },
 324 |   {
 325 |    "cell_type": "code",
 326 |    "execution_count": 7,
 327 |    "metadata": {},
 328 |    "outputs": [
 329 |     {
 330 |      "name": "stdout",
 331 |      "output_type": "stream",
 332 |      "text": [
 333 |       "Pandas(Index=0, Company='Apple', Title='Software Engineer', Job='Software Engineer', Level='Engineer', Date='Aug 7, 2019', Upvotes=0, Offer='No offer', Experience=0.0, Difficulty='Medium', Review='Application  I applied through a staffing agency. I interviewed at Apple (Sunnyvale, CA) in March 2019.  Interview  The interviewer asked me about my background. Asked few questions from the resume. Asked about my proficiency on data structures. Asked me how do you sort hashmap keys based on values.  Interview Questions Write a program that uses two threads to print the numbers from 1 to n.')\n"
 334 |      ]
 335 |     }
 336 |    ],
 337 |    "source": [
 338 |     "# This is what a row looks like as a namedtuple.\n",
 339 |     "for row in df.itertuples():\n",
 340 |     "    print(row)\n",
 341 |     "    break"
 342 |    ]
 343 |   },
 344 |   {
 345 |    "cell_type": "markdown",
 346 |    "metadata": {},
 347 |    "source": [
 348 |     "#### 1.1.3 Converting DataFrame to row-major order\n",
 349 |     "If you need to do a lot of row operations, you might want to convert your `DataFrame` to a NumPy's row-major `ndarray`, then iterating through the rows."
 350 |    ]
 351 |   },
 352 |   {
 353 |    "cell_type": "code",
 354 |    "execution_count": 8,
 355 |    "metadata": {},
 356 |    "outputs": [
 357 |     {
 358 |      "name": "stdout",
 359 |      "output_type": "stream",
 360 |      "text": [
 361 |       "4.55 ms ± 280 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
 362 |      ]
 363 |     }
 364 |    ],
 365 |    "source": [
 366 |     "# Now, iterating through our DataFrame is 100x faster.\n",
 367 |     "%timeit -n1 df_np = df.to_numpy(); rows = [row for row in df_np]"
 368 |    ]
 369 |   },
 370 |   {
 371 |    "cell_type": "markdown",
 372 |    "metadata": {},
 373 |    "source": [
 374 |     "Accessing a row or a column of our `ndarray` takes nanoseconds instead of microseconds."
 375 |    ]
 376 |   },
 377 |   {
 378 |    "cell_type": "code",
 379 |    "execution_count": 9,
 380 |    "metadata": {},
 381 |    "outputs": [
 382 |     {
 383 |      "name": "stdout",
 384 |      "output_type": "stream",
 385 |      "text": [
 386 |       "147 ns ± 1.54 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
 387 |       "204 ns ± 0.678 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n"
 388 |      ]
 389 |     }
 390 |    ],
 391 |    "source": [
 392 |     "df_np = df.to_numpy()\n",
 393 |     "%timeit -n1000 df_np[0]\n",
 394 |     "%timeit -n1000 df_np[:,0]"
 395 |    ]
 396 |   },
 397 |   {
 398 |    "cell_type": "markdown",
 399 |    "metadata": {},
 400 |    "source": [
 401 |     "### 1.2. Ordering slicing operations\n",
 402 |     "Because pandas is column-major, if you want to do multiple slicing operations, always do the column-based slicing operations first.\n",
 403 |     "\n",
 404 |     "For example, if you want to get the review from the first row of the data, there are two slicing operations:\n",
 405 |     "- get row (row-based operation)\n",
 406 |     "- get review (column-based operation)\n",
 407 |     "\n",
 408 |     "Get row -> get review is 25x slower than get review -> get row.\n",
 409 |     "\n",
 410 |     "**Note**: You can also just use `df.loc[0, \"Review\"]` to calculate the memory address to retrieve the item. Its performance is comparable to get review then get row."
 411 |    ]
 412 |   },
 413 |   {
 414 |    "cell_type": "code",
 415 |    "execution_count": 10,
 416 |    "metadata": {},
 417 |    "outputs": [
 418 |     {
 419 |      "name": "stdout",
 420 |      "output_type": "stream",
 421 |      "text": [
 422 |       "5.55 µs ± 1.05 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
 423 |       "136 µs ± 2.57 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n",
 424 |       "6.23 µs ± 264 ns per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n"
 425 |      ]
 426 |     }
 427 |    ],
 428 |    "source": [
 429 |     "%timeit -n1000 df[\"Review\"][0]\n",
 430 |     "%timeit -n1000 df.iloc[0][\"Review\"]\n",
 431 |     "%timeit -n1000 df.loc[0, \"Review\"]"
 432 |    ]
 433 |   },
 434 |   {
 435 |    "cell_type": "markdown",
 436 |    "metadata": {},
 437 |    "source": [
 438 |     "## 2. SettingWithCopyWarning\n",
 439 |     "Sometimes, when you try to assign values to a subset of data in a DataFrame, you get `SettingWithCopyWarning`. Don't ignore the warning because it means sometimes, the assignment works (example 1), but sometimes, it doesn't (example 2)."
 440 |    ]
 441 |   },
 442 |   {
 443 |    "cell_type": "code",
 444 |    "execution_count": 11,
 445 |    "metadata": {},
 446 |    "outputs": [
 447 |     {
 448 |      "name": "stderr",
 449 |      "output_type": "stream",
 450 |      "text": [
 451 |       "/Users/chip/miniconda3/envs/py3/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
 452 |       "A value is trying to be set on a copy of a slice from a DataFrame\n",
 453 |       "\n",
 454 |       "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
 455 |       "  \n"
 456 |      ]
 457 |     },
 458 |     {
 459 |      "data": {
 460 |       "text/html": [
 461 |        "<div>\n",
 462 |        "<style scoped>\n",
 463 |        "    .dataframe tbody tr th:only-of-type {\n",
 464 |        "        vertical-align: middle;\n",
 465 |        "    }\n",
 466 |        "\n",
 467 |        "    .dataframe tbody tr th {\n",
 468 |        "        vertical-align: top;\n",
 469 |        "    }\n",
 470 |        "\n",
 471 |        "    .dataframe thead th {\n",
 472 |        "        text-align: right;\n",
 473 |        "    }\n",
 474 |        "</style>\n",
 475 |        "<table border=\"1\" class=\"dataframe\">\n",
 476 |        "  <thead>\n",
 477 |        "    <tr style=\"text-align: right;\">\n",
 478 |        "      <th></th>\n",
 479 |        "      <th>Company</th>\n",
 480 |        "      <th>Title</th>\n",
 481 |        "      <th>Job</th>\n",
 482 |        "      <th>Level</th>\n",
 483 |        "      <th>Date</th>\n",
 484 |        "      <th>Upvotes</th>\n",
 485 |        "      <th>Offer</th>\n",
 486 |        "      <th>Experience</th>\n",
 487 |        "      <th>Difficulty</th>\n",
 488 |        "      <th>Review</th>\n",
 489 |        "    </tr>\n",
 490 |        "  </thead>\n",
 491 |        "  <tbody>\n",
 492 |        "    <tr>\n",
 493 |        "      <th>0</th>\n",
 494 |        "      <td>Apple</td>\n",
 495 |        "      <td>Software Engineer</td>\n",
 496 |        "      <td>Software Engineer</td>\n",
 497 |        "      <td>Engineer</td>\n",
 498 |        "      <td>Aug 7, 2019</td>\n",
 499 |        "      <td>0</td>\n",
 500 |        "      <td>No offer</td>\n",
 501 |        "      <td>0.0</td>\n",
 502 |        "      <td>Medium</td>\n",
 503 |        "      <td>I like Orange better.</td>\n",
 504 |        "    </tr>\n",
 505 |        "  </tbody>\n",
 506 |        "</table>\n",
 507 |        "</div>"
 508 |       ],
 509 |       "text/plain": [
 510 |        "  Company              Title                Job     Level         Date  \\\n",
 511 |        "0   Apple  Software Engineer  Software Engineer  Engineer  Aug 7, 2019   \n",
 512 |        "\n",
 513 |        "   Upvotes     Offer  Experience Difficulty                 Review  \n",
 514 |        "0        0  No offer         0.0     Medium  I like Orange better.  "
 515 |       ]
 516 |      },
 517 |      "execution_count": 11,
 518 |      "metadata": {},
 519 |      "output_type": "execute_result"
 520 |     }
 521 |    ],
 522 |    "source": [
 523 |     "# Example 1: Changing the review of the first row.\n",
 524 |     "df[\"Review\"][0] = \"I like Orange better.\"\n",
 525 |     "# Even though with the warning, the assignment works. The review is updated.\n",
 526 |     "df.head(1)"
 527 |    ]
 528 |   },
 529 |   {
 530 |    "cell_type": "code",
 531 |    "execution_count": 12,
 532 |    "metadata": {},
 533 |    "outputs": [
 534 |     {
 535 |      "name": "stderr",
 536 |      "output_type": "stream",
 537 |      "text": [
 538 |       "/Users/chip/miniconda3/envs/py3/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
 539 |       "A value is trying to be set on a copy of a slice from a DataFrame.\n",
 540 |       "Try using .loc[row_indexer,col_indexer] = value instead\n",
 541 |       "\n",
 542 |       "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
 543 |       "  \n"
 544 |      ]
 545 |     },
 546 |     {
 547 |      "data": {
 548 |       "text/html": [
 549 |        "<div>\n",
 550 |        "<style scoped>\n",
 551 |        "    .dataframe tbody tr th:only-of-type {\n",
 552 |        "        vertical-align: middle;\n",
 553 |        "    }\n",
 554 |        "\n",
 555 |        "    .dataframe tbody tr th {\n",
 556 |        "        vertical-align: top;\n",
 557 |        "    }\n",
 558 |        "\n",
 559 |        "    .dataframe thead th {\n",
 560 |        "        text-align: right;\n",
 561 |        "    }\n",
 562 |        "</style>\n",
 563 |        "<table border=\"1\" class=\"dataframe\">\n",
 564 |        "  <thead>\n",
 565 |        "    <tr style=\"text-align: right;\">\n",
 566 |        "      <th></th>\n",
 567 |        "      <th>Company</th>\n",
 568 |        "      <th>Title</th>\n",
 569 |        "      <th>Job</th>\n",
 570 |        "      <th>Level</th>\n",
 571 |        "      <th>Date</th>\n",
 572 |        "      <th>Upvotes</th>\n",
 573 |        "      <th>Offer</th>\n",
 574 |        "      <th>Experience</th>\n",
 575 |        "      <th>Difficulty</th>\n",
 576 |        "      <th>Review</th>\n",
 577 |        "    </tr>\n",
 578 |        "  </thead>\n",
 579 |        "  <tbody>\n",
 580 |        "    <tr>\n",
 581 |        "      <th>0</th>\n",
 582 |        "      <td>Apple</td>\n",
 583 |        "      <td>Software Engineer</td>\n",
 584 |        "      <td>Software Engineer</td>\n",
 585 |        "      <td>Engineer</td>\n",
 586 |        "      <td>Aug 7, 2019</td>\n",
 587 |        "      <td>0</td>\n",
 588 |        "      <td>No offer</td>\n",
 589 |        "      <td>0.0</td>\n",
 590 |        "      <td>Medium</td>\n",
 591 |        "      <td>I like Orange better.</td>\n",
 592 |        "    </tr>\n",
 593 |        "  </tbody>\n",
 594 |        "</table>\n",
 595 |        "</div>"
 596 |       ],
 597 |       "text/plain": [
 598 |        "  Company              Title                Job     Level         Date  \\\n",
 599 |        "0   Apple  Software Engineer  Software Engineer  Engineer  Aug 7, 2019   \n",
 600 |        "\n",
 601 |        "   Upvotes     Offer  Experience Difficulty                 Review  \n",
 602 |        "0        0  No offer         0.0     Medium  I like Orange better.  "
 603 |       ]
 604 |      },
 605 |      "execution_count": 12,
 606 |      "metadata": {},
 607 |      "output_type": "execute_result"
 608 |     }
 609 |    ],
 610 |    "source": [
 611 |     "# Example 2: Changing the company name Apple to Orange.\n",
 612 |     "df[df[\"Company\"] == \"Apple\"][\"Company\"] = \"Orange\"\n",
 613 |     "# With the warning, the assignment doesn't work. The company name is still Apple.\n",
 614 |     "df.head(1)"
 615 |    ]
 616 |   },
 617 |   {
 618 |    "cell_type": "markdown",
 619 |    "metadata": {},
 620 |    "source": [
 621 |     "### 2.1. `View` vs. `Copy`\n",
 622 |     "To understand this weird behavior, we need to understand two concepts in pandas: `View` vs. `Copy`.\n",
 623 |     "- `View` is the actual `DataFrame` you want to work with.\n",
 624 |     "- `Copy` is a copy of that actual `DataFrame`, which will be thrown away as soon as the operation is done.\n",
 625 |     "\n",
 626 |     "So if you try to do an assignment on a `Copy`, the assignment won't work. \n",
 627 |     "\n",
 628 |     "`SettingWithCopyWarning` doesn't mean you're making changes to a `Copy`. It means that the thing you're making changes to might be a `Copy` or a `View`, and pandas can't tell you.\n",
 629 |     "\n",
 630 |     "The ambiguity happens because of `__getitem__` operation.\n",
 631 |     "`__getitem__` sometimes returns a `Copy`, sometimes a `View`, and pandas makes no guarantee."
 632 |    ]
 633 |   },
 634 |   {
 635 |    "cell_type": "code",
 636 |    "execution_count": 13,
 637 |    "metadata": {},
 638 |    "outputs": [],
 639 |    "source": [
 640 |     "# df[\"Review\"][0] = \"I like Orange better.\"\n",
 641 |     "# can be understood as\n",
 642 |     "# `df.__getitem__(\"Review\").__setitem__(0, \"I like Orange better.\")`"
 643 |    ]
 644 |   },
 645 |   {
 646 |    "cell_type": "code",
 647 |    "execution_count": 14,
 648 |    "metadata": {},
 649 |    "outputs": [],
 650 |    "source": [
 651 |     "# df[df[\"Company\"] == \"Apple\"][\"Company\"] = \"Orange\"\n",
 652 |     "# can be understood as\n",
 653 |     "# df.__getitem__(where df[\"Company\"] == \"Apple\").__setitem__(\"Company\", \"Orange\")"
 654 |    ]
 655 |   },
 656 |   {
 657 |    "cell_type": "markdown",
 658 |    "metadata": {},
 659 |    "source": [
 660 |     "### 2.2 Solutions\n",
 661 |     "#### 2.2.1 Combine all chained operations into one single operation\n",
 662 |     "To avoid `__getitem__` ambiguity, you can combine all your operations into one single operation.\n",
 663 |     "`.loc[]` is usually great for that."
 664 |    ]
 665 |   },
 666 |   {
 667 |    "cell_type": "code",
 668 |    "execution_count": 15,
 669 |    "metadata": {},
 670 |    "outputs": [
 671 |     {
 672 |      "data": {
 673 |       "text/html": [
 674 |        "<div>\n",
 675 |        "<style scoped>\n",
 676 |        "    .dataframe tbody tr th:only-of-type {\n",
 677 |        "        vertical-align: middle;\n",
 678 |        "    }\n",
 679 |        "\n",
 680 |        "    .dataframe tbody tr th {\n",
 681 |        "        vertical-align: top;\n",
 682 |        "    }\n",
 683 |        "\n",
 684 |        "    .dataframe thead th {\n",
 685 |        "        text-align: right;\n",
 686 |        "    }\n",
 687 |        "</style>\n",
 688 |        "<table border=\"1\" class=\"dataframe\">\n",
 689 |        "  <thead>\n",
 690 |        "    <tr style=\"text-align: right;\">\n",
 691 |        "      <th></th>\n",
 692 |        "      <th>Company</th>\n",
 693 |        "      <th>Title</th>\n",
 694 |        "      <th>Job</th>\n",
 695 |        "      <th>Level</th>\n",
 696 |        "      <th>Date</th>\n",
 697 |        "      <th>Upvotes</th>\n",
 698 |        "      <th>Offer</th>\n",
 699 |        "      <th>Experience</th>\n",
 700 |        "      <th>Difficulty</th>\n",
 701 |        "      <th>Review</th>\n",
 702 |        "    </tr>\n",
 703 |        "  </thead>\n",
 704 |        "  <tbody>\n",
 705 |        "    <tr>\n",
 706 |        "      <th>0</th>\n",
 707 |        "      <td>Apple</td>\n",
 708 |        "      <td>Software Engineer</td>\n",
 709 |        "      <td>Software Engineer</td>\n",
 710 |        "      <td>Engineer</td>\n",
 711 |        "      <td>Aug 7, 2019</td>\n",
 712 |        "      <td>0</td>\n",
 713 |        "      <td>No offer</td>\n",
 714 |        "      <td>0.0</td>\n",
 715 |        "      <td>Medium</td>\n",
 716 |        "      <td>Orange is love. Orange is life.</td>\n",
 717 |        "    </tr>\n",
 718 |        "    <tr>\n",
 719 |        "      <th>1</th>\n",
 720 |        "      <td>Apple</td>\n",
 721 |        "      <td>Software Engineer</td>\n",
 722 |        "      <td>Software Engineer</td>\n",
 723 |        "      <td>Engineer</td>\n",
 724 |        "      <td>Aug 8, 2019</td>\n",
 725 |        "      <td>0</td>\n",
 726 |        "      <td>Accepted offer</td>\n",
 727 |        "      <td>1.0</td>\n",
 728 |        "      <td>Hard</td>\n",
 729 |        "      <td>Application  I applied online. The process too...</td>\n",
 730 |        "    </tr>\n",
 731 |        "    <tr>\n",
 732 |        "      <th>2</th>\n",
 733 |        "      <td>Apple</td>\n",
 734 |        "      <td>Software Engineer</td>\n",
 735 |        "      <td>Software Engineer</td>\n",
 736 |        "      <td>Engineer</td>\n",
 737 |        "      <td>NaN</td>\n",
 738 |        "      <td>0</td>\n",
 739 |        "      <td>Declined offer</td>\n",
 740 |        "      <td>0.0</td>\n",
 741 |        "      <td>Medium</td>\n",
 742 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
 743 |        "    </tr>\n",
 744 |        "    <tr>\n",
 745 |        "      <th>3</th>\n",
 746 |        "      <td>Apple</td>\n",
 747 |        "      <td>Software Engineer</td>\n",
 748 |        "      <td>Software Engineer</td>\n",
 749 |        "      <td>Engineer</td>\n",
 750 |        "      <td>NaN</td>\n",
 751 |        "      <td>9</td>\n",
 752 |        "      <td>Declined offer</td>\n",
 753 |        "      <td>-1.0</td>\n",
 754 |        "      <td>Medium</td>\n",
 755 |        "      <td>Application  The process took a week. I interv...</td>\n",
 756 |        "    </tr>\n",
 757 |        "    <tr>\n",
 758 |        "      <th>4</th>\n",
 759 |        "      <td>Apple</td>\n",
 760 |        "      <td>Software Engineer</td>\n",
 761 |        "      <td>Software Engineer</td>\n",
 762 |        "      <td>Engineer</td>\n",
 763 |        "      <td>May 29, 2009</td>\n",
 764 |        "      <td>2</td>\n",
 765 |        "      <td>No offer</td>\n",
 766 |        "      <td>0.0</td>\n",
 767 |        "      <td>Medium</td>\n",
 768 |        "      <td>Application  I applied through an employee ref...</td>\n",
 769 |        "    </tr>\n",
 770 |        "  </tbody>\n",
 771 |        "</table>\n",
 772 |        "</div>"
 773 |       ],
 774 |       "text/plain": [
 775 |        "  Company              Title                Job     Level          Date  \\\n",
 776 |        "0   Apple  Software Engineer  Software Engineer  Engineer   Aug 7, 2019   \n",
 777 |        "1   Apple  Software Engineer  Software Engineer  Engineer   Aug 8, 2019   \n",
 778 |        "2   Apple  Software Engineer  Software Engineer  Engineer           NaN   \n",
 779 |        "3   Apple  Software Engineer  Software Engineer  Engineer           NaN   \n",
 780 |        "4   Apple  Software Engineer  Software Engineer  Engineer  May 29, 2009   \n",
 781 |        "\n",
 782 |        "   Upvotes           Offer  Experience Difficulty  \\\n",
 783 |        "0        0        No offer         0.0     Medium   \n",
 784 |        "1        0  Accepted offer         1.0       Hard   \n",
 785 |        "2        0  Declined offer         0.0     Medium   \n",
 786 |        "3        9  Declined offer        -1.0     Medium   \n",
 787 |        "4        2        No offer         0.0     Medium   \n",
 788 |        "\n",
 789 |        "                                              Review  \n",
 790 |        "0                    Orange is love. Orange is life.  \n",
 791 |        "1  Application  I applied online. The process too...  \n",
 792 |        "2  Application  The process took 4 weeks. I inter...  \n",
 793 |        "3  Application  The process took a week. I interv...  \n",
 794 |        "4  Application  I applied through an employee ref...  "
 795 |       ]
 796 |      },
 797 |      "execution_count": 15,
 798 |      "metadata": {},
 799 |      "output_type": "execute_result"
 800 |     }
 801 |    ],
 802 |    "source": [
 803 |     "# Changing the review of the first row.\n",
 804 |     "df.loc[0, \"Review\"] = \"Orange is love. Orange is life.\"\n",
 805 |     "df.head()"
 806 |    ]
 807 |   },
 808 |   {
 809 |    "cell_type": "code",
 810 |    "execution_count": 16,
 811 |    "metadata": {},
 812 |    "outputs": [
 813 |     {
 814 |      "data": {
 815 |       "text/html": [
 816 |        "<div>\n",
 817 |        "<style scoped>\n",
 818 |        "    .dataframe tbody tr th:only-of-type {\n",
 819 |        "        vertical-align: middle;\n",
 820 |        "    }\n",
 821 |        "\n",
 822 |        "    .dataframe tbody tr th {\n",
 823 |        "        vertical-align: top;\n",
 824 |        "    }\n",
 825 |        "\n",
 826 |        "    .dataframe thead th {\n",
 827 |        "        text-align: right;\n",
 828 |        "    }\n",
 829 |        "</style>\n",
 830 |        "<table border=\"1\" class=\"dataframe\">\n",
 831 |        "  <thead>\n",
 832 |        "    <tr style=\"text-align: right;\">\n",
 833 |        "      <th></th>\n",
 834 |        "      <th>Company</th>\n",
 835 |        "      <th>Title</th>\n",
 836 |        "      <th>Job</th>\n",
 837 |        "      <th>Level</th>\n",
 838 |        "      <th>Date</th>\n",
 839 |        "      <th>Upvotes</th>\n",
 840 |        "      <th>Offer</th>\n",
 841 |        "      <th>Experience</th>\n",
 842 |        "      <th>Difficulty</th>\n",
 843 |        "      <th>Review</th>\n",
 844 |        "    </tr>\n",
 845 |        "  </thead>\n",
 846 |        "  <tbody>\n",
 847 |        "    <tr>\n",
 848 |        "      <th>0</th>\n",
 849 |        "      <td>Orange</td>\n",
 850 |        "      <td>Software Engineer</td>\n",
 851 |        "      <td>Software Engineer</td>\n",
 852 |        "      <td>Engineer</td>\n",
 853 |        "      <td>Aug 7, 2019</td>\n",
 854 |        "      <td>0</td>\n",
 855 |        "      <td>No offer</td>\n",
 856 |        "      <td>0.0</td>\n",
 857 |        "      <td>Medium</td>\n",
 858 |        "      <td>Orange is love. Orange is life.</td>\n",
 859 |        "    </tr>\n",
 860 |        "    <tr>\n",
 861 |        "      <th>1</th>\n",
 862 |        "      <td>Orange</td>\n",
 863 |        "      <td>Software Engineer</td>\n",
 864 |        "      <td>Software Engineer</td>\n",
 865 |        "      <td>Engineer</td>\n",
 866 |        "      <td>Aug 8, 2019</td>\n",
 867 |        "      <td>0</td>\n",
 868 |        "      <td>Accepted offer</td>\n",
 869 |        "      <td>1.0</td>\n",
 870 |        "      <td>Hard</td>\n",
 871 |        "      <td>Application  I applied online. The process too...</td>\n",
 872 |        "    </tr>\n",
 873 |        "    <tr>\n",
 874 |        "      <th>2</th>\n",
 875 |        "      <td>Orange</td>\n",
 876 |        "      <td>Software Engineer</td>\n",
 877 |        "      <td>Software Engineer</td>\n",
 878 |        "      <td>Engineer</td>\n",
 879 |        "      <td>NaN</td>\n",
 880 |        "      <td>0</td>\n",
 881 |        "      <td>Declined offer</td>\n",
 882 |        "      <td>0.0</td>\n",
 883 |        "      <td>Medium</td>\n",
 884 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
 885 |        "    </tr>\n",
 886 |        "    <tr>\n",
 887 |        "      <th>3</th>\n",
 888 |        "      <td>Orange</td>\n",
 889 |        "      <td>Software Engineer</td>\n",
 890 |        "      <td>Software Engineer</td>\n",
 891 |        "      <td>Engineer</td>\n",
 892 |        "      <td>NaN</td>\n",
 893 |        "      <td>9</td>\n",
 894 |        "      <td>Declined offer</td>\n",
 895 |        "      <td>-1.0</td>\n",
 896 |        "      <td>Medium</td>\n",
 897 |        "      <td>Application  The process took a week. I interv...</td>\n",
 898 |        "    </tr>\n",
 899 |        "    <tr>\n",
 900 |        "      <th>4</th>\n",
 901 |        "      <td>Orange</td>\n",
 902 |        "      <td>Software Engineer</td>\n",
 903 |        "      <td>Software Engineer</td>\n",
 904 |        "      <td>Engineer</td>\n",
 905 |        "      <td>May 29, 2009</td>\n",
 906 |        "      <td>2</td>\n",
 907 |        "      <td>No offer</td>\n",
 908 |        "      <td>0.0</td>\n",
 909 |        "      <td>Medium</td>\n",
 910 |        "      <td>Application  I applied through an employee ref...</td>\n",
 911 |        "    </tr>\n",
 912 |        "  </tbody>\n",
 913 |        "</table>\n",
 914 |        "</div>"
 915 |       ],
 916 |       "text/plain": [
 917 |        "  Company              Title                Job     Level          Date  \\\n",
 918 |        "0  Orange  Software Engineer  Software Engineer  Engineer   Aug 7, 2019   \n",
 919 |        "1  Orange  Software Engineer  Software Engineer  Engineer   Aug 8, 2019   \n",
 920 |        "2  Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
 921 |        "3  Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
 922 |        "4  Orange  Software Engineer  Software Engineer  Engineer  May 29, 2009   \n",
 923 |        "\n",
 924 |        "   Upvotes           Offer  Experience Difficulty  \\\n",
 925 |        "0        0        No offer         0.0     Medium   \n",
 926 |        "1        0  Accepted offer         1.0       Hard   \n",
 927 |        "2        0  Declined offer         0.0     Medium   \n",
 928 |        "3        9  Declined offer        -1.0     Medium   \n",
 929 |        "4        2        No offer         0.0     Medium   \n",
 930 |        "\n",
 931 |        "                                              Review  \n",
 932 |        "0                    Orange is love. Orange is life.  \n",
 933 |        "1  Application  I applied online. The process too...  \n",
 934 |        "2  Application  The process took 4 weeks. I inter...  \n",
 935 |        "3  Application  The process took a week. I interv...  \n",
 936 |        "4  Application  I applied through an employee ref...  "
 937 |       ]
 938 |      },
 939 |      "execution_count": 16,
 940 |      "metadata": {},
 941 |      "output_type": "execute_result"
 942 |     }
 943 |    ],
 944 |    "source": [
 945 |     "# Changing the company name Apple to Orange.\n",
 946 |     "df.loc[df[\"Company\"] == \"Apple\", \"Company\"] = \"Orange\"\n",
 947 |     "df.head()"
 948 |    ]
 949 |   },
 950 |   {
 951 |    "cell_type": "markdown",
 952 |    "metadata": {},
 953 |    "source": [
 954 |     "#### 2.2.2 Raise an error\n",
 955 |     "I believe `SettingWithCopyWarning` should be an Exception instead of a warning. You can change this warning into an exception with pandas' magic `set_option()`."
 956 |    ]
 957 |   },
 958 |   {
 959 |    "cell_type": "code",
 960 |    "execution_count": 17,
 961 |    "metadata": {},
 962 |    "outputs": [],
 963 |    "source": [
 964 |     "pd.set_option(\"mode.chained_assignment\", \"raise\")\n",
 965 |     "# Running this will show you an Exception\n",
 966 |     "# df[\"Review\"][0] = \"I like Orange better.\""
 967 |    ]
 968 |   },
 969 |   {
 970 |    "cell_type": "markdown",
 971 |    "metadata": {},
 972 |    "source": [
 973 |     "## 3. Indexing and slicing\n",
 974 |     "\n",
 975 |     "### 3.1 `.iloc[]`: selecting rows based on integer indices\n",
 976 |     "`.iloc[]` lets you select rows by integer indices."
 977 |    ]
 978 |   },
 979 |   {
 980 |    "cell_type": "code",
 981 |    "execution_count": 18,
 982 |    "metadata": {},
 983 |    "outputs": [
 984 |     {
 985 |      "data": {
 986 |       "text/plain": [
 987 |        "Company                                                  Orange\n",
 988 |        "Title                                         Software Engineer\n",
 989 |        "Job                                           Software Engineer\n",
 990 |        "Level                                                  Engineer\n",
 991 |        "Date                                                        NaN\n",
 992 |        "Upvotes                                                       9\n",
 993 |        "Offer                                            Declined offer\n",
 994 |        "Experience                                                   -1\n",
 995 |        "Difficulty                                               Medium\n",
 996 |        "Review        Application  The process took a week. I interv...\n",
 997 |        "Name: 3, dtype: object"
 998 |       ]
 999 |      },
1000 |      "execution_count": 18,
1001 |      "metadata": {},
1002 |      "output_type": "execute_result"
1003 |     }
1004 |    ],
1005 |    "source": [
1006 |     "# Accessing the third row of a `DataFrame`.\n",
1007 |     "df.iloc[3]"
1008 |    ]
1009 |   },
1010 |   {
1011 |    "cell_type": "markdown",
1012 |    "metadata": {},
1013 |    "source": [
1014 |     "Slicing with `.iloc[]` is similar to slicing in Python. If you want a refresh on how slicing in Python works, see [Python-is-cool](https://github.com/chiphuyen/python-is-cool)."
1015 |    ]
1016 |   },
1017 |   {
1018 |    "cell_type": "code",
1019 |    "execution_count": 19,
1020 |    "metadata": {},
1021 |    "outputs": [
1022 |     {
1023 |      "data": {
1024 |       "text/html": [
1025 |        "<div>\n",
1026 |        "<style scoped>\n",
1027 |        "    .dataframe tbody tr th:only-of-type {\n",
1028 |        "        vertical-align: middle;\n",
1029 |        "    }\n",
1030 |        "\n",
1031 |        "    .dataframe tbody tr th {\n",
1032 |        "        vertical-align: top;\n",
1033 |        "    }\n",
1034 |        "\n",
1035 |        "    .dataframe thead th {\n",
1036 |        "        text-align: right;\n",
1037 |        "    }\n",
1038 |        "</style>\n",
1039 |        "<table border=\"1\" class=\"dataframe\">\n",
1040 |        "  <thead>\n",
1041 |        "    <tr style=\"text-align: right;\">\n",
1042 |        "      <th></th>\n",
1043 |        "      <th>Company</th>\n",
1044 |        "      <th>Title</th>\n",
1045 |        "      <th>Job</th>\n",
1046 |        "      <th>Level</th>\n",
1047 |        "      <th>Date</th>\n",
1048 |        "      <th>Upvotes</th>\n",
1049 |        "      <th>Offer</th>\n",
1050 |        "      <th>Experience</th>\n",
1051 |        "      <th>Difficulty</th>\n",
1052 |        "      <th>Review</th>\n",
1053 |        "    </tr>\n",
1054 |        "  </thead>\n",
1055 |        "  <tbody>\n",
1056 |        "    <tr>\n",
1057 |        "      <th>17648</th>\n",
1058 |        "      <td>Tencent</td>\n",
1059 |        "      <td>Software Engineer</td>\n",
1060 |        "      <td>Software Engineer</td>\n",
1061 |        "      <td>Engineer</td>\n",
1062 |        "      <td>Nov 4, 2012</td>\n",
1063 |        "      <td>0</td>\n",
1064 |        "      <td>No offer</td>\n",
1065 |        "      <td>NaN</td>\n",
1066 |        "      <td>NaN</td>\n",
1067 |        "      <td>Application  I applied online. The process too...</td>\n",
1068 |        "    </tr>\n",
1069 |        "    <tr>\n",
1070 |        "      <th>17649</th>\n",
1071 |        "      <td>Tencent</td>\n",
1072 |        "      <td>Software Engineer</td>\n",
1073 |        "      <td>Software Engineer</td>\n",
1074 |        "      <td>Engineer</td>\n",
1075 |        "      <td>May 25, 2012</td>\n",
1076 |        "      <td>0</td>\n",
1077 |        "      <td>Declined offer</td>\n",
1078 |        "      <td>0.0</td>\n",
1079 |        "      <td>Medium</td>\n",
1080 |        "      <td>Application  I applied online. The process too...</td>\n",
1081 |        "    </tr>\n",
1082 |        "    <tr>\n",
1083 |        "      <th>17650</th>\n",
1084 |        "      <td>Tencent</td>\n",
1085 |        "      <td>Software Engineer</td>\n",
1086 |        "      <td>Software Engineer</td>\n",
1087 |        "      <td>Engineer</td>\n",
1088 |        "      <td>Mar 15, 2014</td>\n",
1089 |        "      <td>0</td>\n",
1090 |        "      <td>No offer</td>\n",
1091 |        "      <td>NaN</td>\n",
1092 |        "      <td>NaN</td>\n",
1093 |        "      <td>Application  I applied through college or univ...</td>\n",
1094 |        "    </tr>\n",
1095 |        "    <tr>\n",
1096 |        "      <th>17651</th>\n",
1097 |        "      <td>Tencent</td>\n",
1098 |        "      <td>Software Engineer</td>\n",
1099 |        "      <td>Software Engineer</td>\n",
1100 |        "      <td>Engineer</td>\n",
1101 |        "      <td>Sep 22, 2015</td>\n",
1102 |        "      <td>0</td>\n",
1103 |        "      <td>Accepted offer</td>\n",
1104 |        "      <td>1.0</td>\n",
1105 |        "      <td>Medium</td>\n",
1106 |        "      <td>Application  I applied through college or univ...</td>\n",
1107 |        "    </tr>\n",
1108 |        "    <tr>\n",
1109 |        "      <th>17652</th>\n",
1110 |        "      <td>Tencent</td>\n",
1111 |        "      <td>Software Engineer</td>\n",
1112 |        "      <td>Software Engineer</td>\n",
1113 |        "      <td>Engineer</td>\n",
1114 |        "      <td>Jul 4, 2017</td>\n",
1115 |        "      <td>0</td>\n",
1116 |        "      <td>Declined offer</td>\n",
1117 |        "      <td>1.0</td>\n",
1118 |        "      <td>Medium</td>\n",
1119 |        "      <td>Application  I applied through college or univ...</td>\n",
1120 |        "    </tr>\n",
1121 |        "    <tr>\n",
1122 |        "      <th>17653</th>\n",
1123 |        "      <td>Tencent</td>\n",
1124 |        "      <td>Software Engineer</td>\n",
1125 |        "      <td>Software Engineer</td>\n",
1126 |        "      <td>Engineer</td>\n",
1127 |        "      <td>Sep 30, 2016</td>\n",
1128 |        "      <td>0</td>\n",
1129 |        "      <td>Declined offer</td>\n",
1130 |        "      <td>0.0</td>\n",
1131 |        "      <td>Easy</td>\n",
1132 |        "      <td>Application  I applied online. The process too...</td>\n",
1133 |        "    </tr>\n",
1134 |        "  </tbody>\n",
1135 |        "</table>\n",
1136 |        "</div>"
1137 |       ],
1138 |       "text/plain": [
1139 |        "       Company              Title                Job     Level          Date  \\\n",
1140 |        "17648  Tencent  Software Engineer  Software Engineer  Engineer   Nov 4, 2012   \n",
1141 |        "17649  Tencent  Software Engineer  Software Engineer  Engineer  May 25, 2012   \n",
1142 |        "17650  Tencent  Software Engineer  Software Engineer  Engineer  Mar 15, 2014   \n",
1143 |        "17651  Tencent  Software Engineer  Software Engineer  Engineer  Sep 22, 2015   \n",
1144 |        "17652  Tencent  Software Engineer  Software Engineer  Engineer   Jul 4, 2017   \n",
1145 |        "17653  Tencent  Software Engineer  Software Engineer  Engineer  Sep 30, 2016   \n",
1146 |        "\n",
1147 |        "       Upvotes           Offer  Experience Difficulty  \\\n",
1148 |        "17648        0        No offer         NaN        NaN   \n",
1149 |        "17649        0  Declined offer         0.0     Medium   \n",
1150 |        "17650        0        No offer         NaN        NaN   \n",
1151 |        "17651        0  Accepted offer         1.0     Medium   \n",
1152 |        "17652        0  Declined offer         1.0     Medium   \n",
1153 |        "17653        0  Declined offer         0.0       Easy   \n",
1154 |        "\n",
1155 |        "                                                  Review  \n",
1156 |        "17648  Application  I applied online. The process too...  \n",
1157 |        "17649  Application  I applied online. The process too...  \n",
1158 |        "17650  Application  I applied through college or univ...  \n",
1159 |        "17651  Application  I applied through college or univ...  \n",
1160 |        "17652  Application  I applied through college or univ...  \n",
1161 |        "17653  Application  I applied online. The process too...  "
1162 |       ]
1163 |      },
1164 |      "execution_count": 19,
1165 |      "metadata": {},
1166 |      "output_type": "execute_result"
1167 |     }
1168 |    ],
1169 |    "source": [
1170 |     "# Selecting the last 6 rows\n",
1171 |     "df.iloc[-6:]"
1172 |    ]
1173 |   },
1174 |   {
1175 |    "cell_type": "code",
1176 |    "execution_count": 20,
1177 |    "metadata": {},
1178 |    "outputs": [
1179 |     {
1180 |      "data": {
1181 |       "text/html": [
1182 |        "<div>\n",
1183 |        "<style scoped>\n",
1184 |        "    .dataframe tbody tr th:only-of-type {\n",
1185 |        "        vertical-align: middle;\n",
1186 |        "    }\n",
1187 |        "\n",
1188 |        "    .dataframe tbody tr th {\n",
1189 |        "        vertical-align: top;\n",
1190 |        "    }\n",
1191 |        "\n",
1192 |        "    .dataframe thead th {\n",
1193 |        "        text-align: right;\n",
1194 |        "    }\n",
1195 |        "</style>\n",
1196 |        "<table border=\"1\" class=\"dataframe\">\n",
1197 |        "  <thead>\n",
1198 |        "    <tr style=\"text-align: right;\">\n",
1199 |        "      <th></th>\n",
1200 |        "      <th>Company</th>\n",
1201 |        "      <th>Title</th>\n",
1202 |        "      <th>Job</th>\n",
1203 |        "      <th>Level</th>\n",
1204 |        "      <th>Date</th>\n",
1205 |        "      <th>Upvotes</th>\n",
1206 |        "      <th>Offer</th>\n",
1207 |        "      <th>Experience</th>\n",
1208 |        "      <th>Difficulty</th>\n",
1209 |        "      <th>Review</th>\n",
1210 |        "    </tr>\n",
1211 |        "  </thead>\n",
1212 |        "  <tbody>\n",
1213 |        "    <tr>\n",
1214 |        "      <th>17648</th>\n",
1215 |        "      <td>Tencent</td>\n",
1216 |        "      <td>Software Engineer</td>\n",
1217 |        "      <td>Software Engineer</td>\n",
1218 |        "      <td>Engineer</td>\n",
1219 |        "      <td>Nov 4, 2012</td>\n",
1220 |        "      <td>0</td>\n",
1221 |        "      <td>No offer</td>\n",
1222 |        "      <td>NaN</td>\n",
1223 |        "      <td>NaN</td>\n",
1224 |        "      <td>Application  I applied online. The process too...</td>\n",
1225 |        "    </tr>\n",
1226 |        "    <tr>\n",
1227 |        "      <th>17650</th>\n",
1228 |        "      <td>Tencent</td>\n",
1229 |        "      <td>Software Engineer</td>\n",
1230 |        "      <td>Software Engineer</td>\n",
1231 |        "      <td>Engineer</td>\n",
1232 |        "      <td>Mar 15, 2014</td>\n",
1233 |        "      <td>0</td>\n",
1234 |        "      <td>No offer</td>\n",
1235 |        "      <td>NaN</td>\n",
1236 |        "      <td>NaN</td>\n",
1237 |        "      <td>Application  I applied through college or univ...</td>\n",
1238 |        "    </tr>\n",
1239 |        "    <tr>\n",
1240 |        "      <th>17652</th>\n",
1241 |        "      <td>Tencent</td>\n",
1242 |        "      <td>Software Engineer</td>\n",
1243 |        "      <td>Software Engineer</td>\n",
1244 |        "      <td>Engineer</td>\n",
1245 |        "      <td>Jul 4, 2017</td>\n",
1246 |        "      <td>0</td>\n",
1247 |        "      <td>Declined offer</td>\n",
1248 |        "      <td>1.0</td>\n",
1249 |        "      <td>Medium</td>\n",
1250 |        "      <td>Application  I applied through college or univ...</td>\n",
1251 |        "    </tr>\n",
1252 |        "  </tbody>\n",
1253 |        "</table>\n",
1254 |        "</div>"
1255 |       ],
1256 |       "text/plain": [
1257 |        "       Company              Title                Job     Level          Date  \\\n",
1258 |        "17648  Tencent  Software Engineer  Software Engineer  Engineer   Nov 4, 2012   \n",
1259 |        "17650  Tencent  Software Engineer  Software Engineer  Engineer  Mar 15, 2014   \n",
1260 |        "17652  Tencent  Software Engineer  Software Engineer  Engineer   Jul 4, 2017   \n",
1261 |        "\n",
1262 |        "       Upvotes           Offer  Experience Difficulty  \\\n",
1263 |        "17648        0        No offer         NaN        NaN   \n",
1264 |        "17650        0        No offer         NaN        NaN   \n",
1265 |        "17652        0  Declined offer         1.0     Medium   \n",
1266 |        "\n",
1267 |        "                                                  Review  \n",
1268 |        "17648  Application  I applied online. The process too...  \n",
1269 |        "17650  Application  I applied through college or univ...  \n",
1270 |        "17652  Application  I applied through college or univ...  "
1271 |       ]
1272 |      },
1273 |      "execution_count": 20,
1274 |      "metadata": {},
1275 |      "output_type": "execute_result"
1276 |     }
1277 |    ],
1278 |    "source": [
1279 |     "# Selecting 1 from every 2 rows in the last 6 rows\n",
1280 |     "df.iloc[-6::2]"
1281 |    ]
1282 |   },
1283 |   {
1284 |    "cell_type": "markdown",
1285 |    "metadata": {},
1286 |    "source": [
1287 |     "### 3.2 `.loc[]`: selecting rows by labels or boolean masks\n",
1288 |     "`.loc[]` lets you select rows based on one of the two things:\n",
1289 |     "- boolean masks\n",
1290 |     "- labels\n",
1291 |     "\n",
1292 |     "#### 3.2.1 Selecting rows by boolean masks\n",
1293 |     "If you want to select all the rows where candidates declined offer, you can do it with two steps:\n",
1294 |     "1. Create a boolean mask on whether the \"Offer\" column equals to \"Declined offer\"\n",
1295 |     "2. Use that mask to select rows"
1296 |    ]
1297 |   },
1298 |   {
1299 |    "cell_type": "code",
1300 |    "execution_count": 21,
1301 |    "metadata": {},
1302 |    "outputs": [
1303 |     {
1304 |      "data": {
1305 |       "text/html": [
1306 |        "<div>\n",
1307 |        "<style scoped>\n",
1308 |        "    .dataframe tbody tr th:only-of-type {\n",
1309 |        "        vertical-align: middle;\n",
1310 |        "    }\n",
1311 |        "\n",
1312 |        "    .dataframe tbody tr th {\n",
1313 |        "        vertical-align: top;\n",
1314 |        "    }\n",
1315 |        "\n",
1316 |        "    .dataframe thead th {\n",
1317 |        "        text-align: right;\n",
1318 |        "    }\n",
1319 |        "</style>\n",
1320 |        "<table border=\"1\" class=\"dataframe\">\n",
1321 |        "  <thead>\n",
1322 |        "    <tr style=\"text-align: right;\">\n",
1323 |        "      <th></th>\n",
1324 |        "      <th>Company</th>\n",
1325 |        "      <th>Title</th>\n",
1326 |        "      <th>Job</th>\n",
1327 |        "      <th>Level</th>\n",
1328 |        "      <th>Date</th>\n",
1329 |        "      <th>Upvotes</th>\n",
1330 |        "      <th>Offer</th>\n",
1331 |        "      <th>Experience</th>\n",
1332 |        "      <th>Difficulty</th>\n",
1333 |        "      <th>Review</th>\n",
1334 |        "    </tr>\n",
1335 |        "  </thead>\n",
1336 |        "  <tbody>\n",
1337 |        "    <tr>\n",
1338 |        "      <th>2</th>\n",
1339 |        "      <td>Orange</td>\n",
1340 |        "      <td>Software Engineer</td>\n",
1341 |        "      <td>Software Engineer</td>\n",
1342 |        "      <td>Engineer</td>\n",
1343 |        "      <td>NaN</td>\n",
1344 |        "      <td>0</td>\n",
1345 |        "      <td>Declined offer</td>\n",
1346 |        "      <td>0.0</td>\n",
1347 |        "      <td>Medium</td>\n",
1348 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
1349 |        "    </tr>\n",
1350 |        "    <tr>\n",
1351 |        "      <th>3</th>\n",
1352 |        "      <td>Orange</td>\n",
1353 |        "      <td>Software Engineer</td>\n",
1354 |        "      <td>Software Engineer</td>\n",
1355 |        "      <td>Engineer</td>\n",
1356 |        "      <td>NaN</td>\n",
1357 |        "      <td>9</td>\n",
1358 |        "      <td>Declined offer</td>\n",
1359 |        "      <td>-1.0</td>\n",
1360 |        "      <td>Medium</td>\n",
1361 |        "      <td>Application  The process took a week. I interv...</td>\n",
1362 |        "    </tr>\n",
1363 |        "    <tr>\n",
1364 |        "      <th>7</th>\n",
1365 |        "      <td>Orange</td>\n",
1366 |        "      <td>Software Engineer</td>\n",
1367 |        "      <td>Software Engineer</td>\n",
1368 |        "      <td>Engineer</td>\n",
1369 |        "      <td>Jul 26, 2019</td>\n",
1370 |        "      <td>1</td>\n",
1371 |        "      <td>Declined offer</td>\n",
1372 |        "      <td>-1.0</td>\n",
1373 |        "      <td>Medium</td>\n",
1374 |        "      <td>Application  The process took 4+ weeks. I inte...</td>\n",
1375 |        "    </tr>\n",
1376 |        "    <tr>\n",
1377 |        "      <th>17</th>\n",
1378 |        "      <td>Orange</td>\n",
1379 |        "      <td>Software Engineer</td>\n",
1380 |        "      <td>Software Engineer</td>\n",
1381 |        "      <td>Engineer</td>\n",
1382 |        "      <td>Feb 27, 2010</td>\n",
1383 |        "      <td>7</td>\n",
1384 |        "      <td>Declined offer</td>\n",
1385 |        "      <td>-1.0</td>\n",
1386 |        "      <td>Medium</td>\n",
1387 |        "      <td>Application  The process took 1 day. I intervi...</td>\n",
1388 |        "    </tr>\n",
1389 |        "    <tr>\n",
1390 |        "      <th>65</th>\n",
1391 |        "      <td>Orange</td>\n",
1392 |        "      <td>Software Engineer</td>\n",
1393 |        "      <td>Software Engineer</td>\n",
1394 |        "      <td>Engineer</td>\n",
1395 |        "      <td>May 6, 2012</td>\n",
1396 |        "      <td>1</td>\n",
1397 |        "      <td>Declined offer</td>\n",
1398 |        "      <td>1.0</td>\n",
1399 |        "      <td>Easy</td>\n",
1400 |        "      <td>Application  The process took 2 days. I interv...</td>\n",
1401 |        "    </tr>\n",
1402 |        "    <tr>\n",
1403 |        "      <th>...</th>\n",
1404 |        "      <td>...</td>\n",
1405 |        "      <td>...</td>\n",
1406 |        "      <td>...</td>\n",
1407 |        "      <td>...</td>\n",
1408 |        "      <td>...</td>\n",
1409 |        "      <td>...</td>\n",
1410 |        "      <td>...</td>\n",
1411 |        "      <td>...</td>\n",
1412 |        "      <td>...</td>\n",
1413 |        "      <td>...</td>\n",
1414 |        "    </tr>\n",
1415 |        "    <tr>\n",
1416 |        "      <th>17643</th>\n",
1417 |        "      <td>Tencent</td>\n",
1418 |        "      <td>Software Engineer</td>\n",
1419 |        "      <td>Software Engineer</td>\n",
1420 |        "      <td>Engineer</td>\n",
1421 |        "      <td>Apr 9, 2016</td>\n",
1422 |        "      <td>0</td>\n",
1423 |        "      <td>Declined offer</td>\n",
1424 |        "      <td>1.0</td>\n",
1425 |        "      <td>Medium</td>\n",
1426 |        "      <td>Application  I applied online. I interviewed a...</td>\n",
1427 |        "    </tr>\n",
1428 |        "    <tr>\n",
1429 |        "      <th>17646</th>\n",
1430 |        "      <td>Tencent</td>\n",
1431 |        "      <td>Software Engineer</td>\n",
1432 |        "      <td>Software Engineer</td>\n",
1433 |        "      <td>Engineer</td>\n",
1434 |        "      <td>May 28, 2010</td>\n",
1435 |        "      <td>0</td>\n",
1436 |        "      <td>Declined offer</td>\n",
1437 |        "      <td>0.0</td>\n",
1438 |        "      <td>Easy</td>\n",
1439 |        "      <td>Application  I applied through an employee ref...</td>\n",
1440 |        "    </tr>\n",
1441 |        "    <tr>\n",
1442 |        "      <th>17649</th>\n",
1443 |        "      <td>Tencent</td>\n",
1444 |        "      <td>Software Engineer</td>\n",
1445 |        "      <td>Software Engineer</td>\n",
1446 |        "      <td>Engineer</td>\n",
1447 |        "      <td>May 25, 2012</td>\n",
1448 |        "      <td>0</td>\n",
1449 |        "      <td>Declined offer</td>\n",
1450 |        "      <td>0.0</td>\n",
1451 |        "      <td>Medium</td>\n",
1452 |        "      <td>Application  I applied online. The process too...</td>\n",
1453 |        "    </tr>\n",
1454 |        "    <tr>\n",
1455 |        "      <th>17652</th>\n",
1456 |        "      <td>Tencent</td>\n",
1457 |        "      <td>Software Engineer</td>\n",
1458 |        "      <td>Software Engineer</td>\n",
1459 |        "      <td>Engineer</td>\n",
1460 |        "      <td>Jul 4, 2017</td>\n",
1461 |        "      <td>0</td>\n",
1462 |        "      <td>Declined offer</td>\n",
1463 |        "      <td>1.0</td>\n",
1464 |        "      <td>Medium</td>\n",
1465 |        "      <td>Application  I applied through college or univ...</td>\n",
1466 |        "    </tr>\n",
1467 |        "    <tr>\n",
1468 |        "      <th>17653</th>\n",
1469 |        "      <td>Tencent</td>\n",
1470 |        "      <td>Software Engineer</td>\n",
1471 |        "      <td>Software Engineer</td>\n",
1472 |        "      <td>Engineer</td>\n",
1473 |        "      <td>Sep 30, 2016</td>\n",
1474 |        "      <td>0</td>\n",
1475 |        "      <td>Declined offer</td>\n",
1476 |        "      <td>0.0</td>\n",
1477 |        "      <td>Easy</td>\n",
1478 |        "      <td>Application  I applied online. The process too...</td>\n",
1479 |        "    </tr>\n",
1480 |        "  </tbody>\n",
1481 |        "</table>\n",
1482 |        "<p>1135 rows × 10 columns</p>\n",
1483 |        "</div>"
1484 |       ],
1485 |       "text/plain": [
1486 |        "       Company              Title                Job     Level          Date  \\\n",
1487 |        "2       Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
1488 |        "3       Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
1489 |        "7       Orange  Software Engineer  Software Engineer  Engineer  Jul 26, 2019   \n",
1490 |        "17      Orange  Software Engineer  Software Engineer  Engineer  Feb 27, 2010   \n",
1491 |        "65      Orange  Software Engineer  Software Engineer  Engineer   May 6, 2012   \n",
1492 |        "...        ...                ...                ...       ...           ...   \n",
1493 |        "17643  Tencent  Software Engineer  Software Engineer  Engineer   Apr 9, 2016   \n",
1494 |        "17646  Tencent  Software Engineer  Software Engineer  Engineer  May 28, 2010   \n",
1495 |        "17649  Tencent  Software Engineer  Software Engineer  Engineer  May 25, 2012   \n",
1496 |        "17652  Tencent  Software Engineer  Software Engineer  Engineer   Jul 4, 2017   \n",
1497 |        "17653  Tencent  Software Engineer  Software Engineer  Engineer  Sep 30, 2016   \n",
1498 |        "\n",
1499 |        "       Upvotes           Offer  Experience Difficulty  \\\n",
1500 |        "2            0  Declined offer         0.0     Medium   \n",
1501 |        "3            9  Declined offer        -1.0     Medium   \n",
1502 |        "7            1  Declined offer        -1.0     Medium   \n",
1503 |        "17           7  Declined offer        -1.0     Medium   \n",
1504 |        "65           1  Declined offer         1.0       Easy   \n",
1505 |        "...        ...             ...         ...        ...   \n",
1506 |        "17643        0  Declined offer         1.0     Medium   \n",
1507 |        "17646        0  Declined offer         0.0       Easy   \n",
1508 |        "17649        0  Declined offer         0.0     Medium   \n",
1509 |        "17652        0  Declined offer         1.0     Medium   \n",
1510 |        "17653        0  Declined offer         0.0       Easy   \n",
1511 |        "\n",
1512 |        "                                                  Review  \n",
1513 |        "2      Application  The process took 4 weeks. I inter...  \n",
1514 |        "3      Application  The process took a week. I interv...  \n",
1515 |        "7      Application  The process took 4+ weeks. I inte...  \n",
1516 |        "17     Application  The process took 1 day. I intervi...  \n",
1517 |        "65     Application  The process took 2 days. I interv...  \n",
1518 |        "...                                                  ...  \n",
1519 |        "17643  Application  I applied online. I interviewed a...  \n",
1520 |        "17646  Application  I applied through an employee ref...  \n",
1521 |        "17649  Application  I applied online. The process too...  \n",
1522 |        "17652  Application  I applied through college or univ...  \n",
1523 |        "17653  Application  I applied online. The process too...  \n",
1524 |        "\n",
1525 |        "[1135 rows x 10 columns]"
1526 |       ]
1527 |      },
1528 |      "execution_count": 21,
1529 |      "metadata": {},
1530 |      "output_type": "execute_result"
1531 |     }
1532 |    ],
1533 |    "source": [
1534 |     "df.loc[df[\"Offer\"] == \"Declined offer\"]\n",
1535 |     "# This is equivalent to:\n",
1536 |     "# mask = df[\"Offer\"] == \"Declined offer\"\n",
1537 |     "# df.loc[mask]"
1538 |    ]
1539 |   },
1540 |   {
1541 |    "cell_type": "markdown",
1542 |    "metadata": {},
1543 |    "source": [
1544 |     "#### 3.2.2 Selecting rows by labels\n",
1545 |     "##### 3.2.2.1 Creating labels\n",
1546 |     "Currently, our `DataFrame` has no labels yet. To create labels, use `.set_index()`.\n",
1547 |     "\n",
1548 |     "1. Labels can be integers or strings\n",
1549 |     "2. A DataFrame can have multiple labels"
1550 |    ]
1551 |   },
1552 |   {
1553 |    "cell_type": "code",
1554 |    "execution_count": 22,
1555 |    "metadata": {},
1556 |    "outputs": [
1557 |     {
1558 |      "data": {
1559 |       "text/html": [
1560 |        "<div>\n",
1561 |        "<style scoped>\n",
1562 |        "    .dataframe tbody tr th:only-of-type {\n",
1563 |        "        vertical-align: middle;\n",
1564 |        "    }\n",
1565 |        "\n",
1566 |        "    .dataframe tbody tr th {\n",
1567 |        "        vertical-align: top;\n",
1568 |        "    }\n",
1569 |        "\n",
1570 |        "    .dataframe thead th {\n",
1571 |        "        text-align: right;\n",
1572 |        "    }\n",
1573 |        "</style>\n",
1574 |        "<table border=\"1\" class=\"dataframe\">\n",
1575 |        "  <thead>\n",
1576 |        "    <tr style=\"text-align: right;\">\n",
1577 |        "      <th></th>\n",
1578 |        "      <th>Company</th>\n",
1579 |        "      <th>Title</th>\n",
1580 |        "      <th>Job</th>\n",
1581 |        "      <th>Level</th>\n",
1582 |        "      <th>Date</th>\n",
1583 |        "      <th>Upvotes</th>\n",
1584 |        "      <th>Offer</th>\n",
1585 |        "      <th>Experience</th>\n",
1586 |        "      <th>Difficulty</th>\n",
1587 |        "      <th>Review</th>\n",
1588 |        "    </tr>\n",
1589 |        "    <tr>\n",
1590 |        "      <th>Type</th>\n",
1591 |        "      <th></th>\n",
1592 |        "      <th></th>\n",
1593 |        "      <th></th>\n",
1594 |        "      <th></th>\n",
1595 |        "      <th></th>\n",
1596 |        "      <th></th>\n",
1597 |        "      <th></th>\n",
1598 |        "      <th></th>\n",
1599 |        "      <th></th>\n",
1600 |        "      <th></th>\n",
1601 |        "    </tr>\n",
1602 |        "  </thead>\n",
1603 |        "  <tbody>\n",
1604 |        "    <tr>\n",
1605 |        "      <th>Hardware</th>\n",
1606 |        "      <td>Orange</td>\n",
1607 |        "      <td>Software Engineer</td>\n",
1608 |        "      <td>Software Engineer</td>\n",
1609 |        "      <td>Engineer</td>\n",
1610 |        "      <td>Aug 7, 2019</td>\n",
1611 |        "      <td>0</td>\n",
1612 |        "      <td>No offer</td>\n",
1613 |        "      <td>0.0</td>\n",
1614 |        "      <td>Medium</td>\n",
1615 |        "      <td>Orange is love. Orange is life.</td>\n",
1616 |        "    </tr>\n",
1617 |        "    <tr>\n",
1618 |        "      <th>Hardware</th>\n",
1619 |        "      <td>Orange</td>\n",
1620 |        "      <td>Software Engineer</td>\n",
1621 |        "      <td>Software Engineer</td>\n",
1622 |        "      <td>Engineer</td>\n",
1623 |        "      <td>Aug 8, 2019</td>\n",
1624 |        "      <td>0</td>\n",
1625 |        "      <td>Accepted offer</td>\n",
1626 |        "      <td>1.0</td>\n",
1627 |        "      <td>Hard</td>\n",
1628 |        "      <td>Application  I applied online. The process too...</td>\n",
1629 |        "    </tr>\n",
1630 |        "    <tr>\n",
1631 |        "      <th>Hardware</th>\n",
1632 |        "      <td>Orange</td>\n",
1633 |        "      <td>Software Engineer</td>\n",
1634 |        "      <td>Software Engineer</td>\n",
1635 |        "      <td>Engineer</td>\n",
1636 |        "      <td>NaN</td>\n",
1637 |        "      <td>0</td>\n",
1638 |        "      <td>Declined offer</td>\n",
1639 |        "      <td>0.0</td>\n",
1640 |        "      <td>Medium</td>\n",
1641 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
1642 |        "    </tr>\n",
1643 |        "    <tr>\n",
1644 |        "      <th>Hardware</th>\n",
1645 |        "      <td>Orange</td>\n",
1646 |        "      <td>Software Engineer</td>\n",
1647 |        "      <td>Software Engineer</td>\n",
1648 |        "      <td>Engineer</td>\n",
1649 |        "      <td>NaN</td>\n",
1650 |        "      <td>9</td>\n",
1651 |        "      <td>Declined offer</td>\n",
1652 |        "      <td>-1.0</td>\n",
1653 |        "      <td>Medium</td>\n",
1654 |        "      <td>Application  The process took a week. I interv...</td>\n",
1655 |        "    </tr>\n",
1656 |        "    <tr>\n",
1657 |        "      <th>Hardware</th>\n",
1658 |        "      <td>Orange</td>\n",
1659 |        "      <td>Software Engineer</td>\n",
1660 |        "      <td>Software Engineer</td>\n",
1661 |        "      <td>Engineer</td>\n",
1662 |        "      <td>May 29, 2009</td>\n",
1663 |        "      <td>2</td>\n",
1664 |        "      <td>No offer</td>\n",
1665 |        "      <td>0.0</td>\n",
1666 |        "      <td>Medium</td>\n",
1667 |        "      <td>Application  I applied through an employee ref...</td>\n",
1668 |        "    </tr>\n",
1669 |        "    <tr>\n",
1670 |        "      <th>...</th>\n",
1671 |        "      <td>...</td>\n",
1672 |        "      <td>...</td>\n",
1673 |        "      <td>...</td>\n",
1674 |        "      <td>...</td>\n",
1675 |        "      <td>...</td>\n",
1676 |        "      <td>...</td>\n",
1677 |        "      <td>...</td>\n",
1678 |        "      <td>...</td>\n",
1679 |        "      <td>...</td>\n",
1680 |        "      <td>...</td>\n",
1681 |        "    </tr>\n",
1682 |        "    <tr>\n",
1683 |        "      <th>Software</th>\n",
1684 |        "      <td>Tencent</td>\n",
1685 |        "      <td>Software Engineer</td>\n",
1686 |        "      <td>Software Engineer</td>\n",
1687 |        "      <td>Engineer</td>\n",
1688 |        "      <td>May 25, 2012</td>\n",
1689 |        "      <td>0</td>\n",
1690 |        "      <td>Declined offer</td>\n",
1691 |        "      <td>0.0</td>\n",
1692 |        "      <td>Medium</td>\n",
1693 |        "      <td>Application  I applied online. The process too...</td>\n",
1694 |        "    </tr>\n",
1695 |        "    <tr>\n",
1696 |        "      <th>Software</th>\n",
1697 |        "      <td>Tencent</td>\n",
1698 |        "      <td>Software Engineer</td>\n",
1699 |        "      <td>Software Engineer</td>\n",
1700 |        "      <td>Engineer</td>\n",
1701 |        "      <td>Mar 15, 2014</td>\n",
1702 |        "      <td>0</td>\n",
1703 |        "      <td>No offer</td>\n",
1704 |        "      <td>NaN</td>\n",
1705 |        "      <td>NaN</td>\n",
1706 |        "      <td>Application  I applied through college or univ...</td>\n",
1707 |        "    </tr>\n",
1708 |        "    <tr>\n",
1709 |        "      <th>Software</th>\n",
1710 |        "      <td>Tencent</td>\n",
1711 |        "      <td>Software Engineer</td>\n",
1712 |        "      <td>Software Engineer</td>\n",
1713 |        "      <td>Engineer</td>\n",
1714 |        "      <td>Sep 22, 2015</td>\n",
1715 |        "      <td>0</td>\n",
1716 |        "      <td>Accepted offer</td>\n",
1717 |        "      <td>1.0</td>\n",
1718 |        "      <td>Medium</td>\n",
1719 |        "      <td>Application  I applied through college or univ...</td>\n",
1720 |        "    </tr>\n",
1721 |        "    <tr>\n",
1722 |        "      <th>Software</th>\n",
1723 |        "      <td>Tencent</td>\n",
1724 |        "      <td>Software Engineer</td>\n",
1725 |        "      <td>Software Engineer</td>\n",
1726 |        "      <td>Engineer</td>\n",
1727 |        "      <td>Jul 4, 2017</td>\n",
1728 |        "      <td>0</td>\n",
1729 |        "      <td>Declined offer</td>\n",
1730 |        "      <td>1.0</td>\n",
1731 |        "      <td>Medium</td>\n",
1732 |        "      <td>Application  I applied through college or univ...</td>\n",
1733 |        "    </tr>\n",
1734 |        "    <tr>\n",
1735 |        "      <th>Software</th>\n",
1736 |        "      <td>Tencent</td>\n",
1737 |        "      <td>Software Engineer</td>\n",
1738 |        "      <td>Software Engineer</td>\n",
1739 |        "      <td>Engineer</td>\n",
1740 |        "      <td>Sep 30, 2016</td>\n",
1741 |        "      <td>0</td>\n",
1742 |        "      <td>Declined offer</td>\n",
1743 |        "      <td>0.0</td>\n",
1744 |        "      <td>Easy</td>\n",
1745 |        "      <td>Application  I applied online. The process too...</td>\n",
1746 |        "    </tr>\n",
1747 |        "  </tbody>\n",
1748 |        "</table>\n",
1749 |        "<p>17654 rows × 10 columns</p>\n",
1750 |        "</div>"
1751 |       ],
1752 |       "text/plain": [
1753 |        "          Company              Title                Job     Level  \\\n",
1754 |        "Type                                                                \n",
1755 |        "Hardware   Orange  Software Engineer  Software Engineer  Engineer   \n",
1756 |        "Hardware   Orange  Software Engineer  Software Engineer  Engineer   \n",
1757 |        "Hardware   Orange  Software Engineer  Software Engineer  Engineer   \n",
1758 |        "Hardware   Orange  Software Engineer  Software Engineer  Engineer   \n",
1759 |        "Hardware   Orange  Software Engineer  Software Engineer  Engineer   \n",
1760 |        "...           ...                ...                ...       ...   \n",
1761 |        "Software  Tencent  Software Engineer  Software Engineer  Engineer   \n",
1762 |        "Software  Tencent  Software Engineer  Software Engineer  Engineer   \n",
1763 |        "Software  Tencent  Software Engineer  Software Engineer  Engineer   \n",
1764 |        "Software  Tencent  Software Engineer  Software Engineer  Engineer   \n",
1765 |        "Software  Tencent  Software Engineer  Software Engineer  Engineer   \n",
1766 |        "\n",
1767 |        "                  Date  Upvotes           Offer  Experience Difficulty  \\\n",
1768 |        "Type                                                                     \n",
1769 |        "Hardware   Aug 7, 2019        0        No offer         0.0     Medium   \n",
1770 |        "Hardware   Aug 8, 2019        0  Accepted offer         1.0       Hard   \n",
1771 |        "Hardware           NaN        0  Declined offer         0.0     Medium   \n",
1772 |        "Hardware           NaN        9  Declined offer        -1.0     Medium   \n",
1773 |        "Hardware  May 29, 2009        2        No offer         0.0     Medium   \n",
1774 |        "...                ...      ...             ...         ...        ...   \n",
1775 |        "Software  May 25, 2012        0  Declined offer         0.0     Medium   \n",
1776 |        "Software  Mar 15, 2014        0        No offer         NaN        NaN   \n",
1777 |        "Software  Sep 22, 2015        0  Accepted offer         1.0     Medium   \n",
1778 |        "Software   Jul 4, 2017        0  Declined offer         1.0     Medium   \n",
1779 |        "Software  Sep 30, 2016        0  Declined offer         0.0       Easy   \n",
1780 |        "\n",
1781 |        "                                                     Review  \n",
1782 |        "Type                                                         \n",
1783 |        "Hardware                    Orange is love. Orange is life.  \n",
1784 |        "Hardware  Application  I applied online. The process too...  \n",
1785 |        "Hardware  Application  The process took 4 weeks. I inter...  \n",
1786 |        "Hardware  Application  The process took a week. I interv...  \n",
1787 |        "Hardware  Application  I applied through an employee ref...  \n",
1788 |        "...                                                     ...  \n",
1789 |        "Software  Application  I applied online. The process too...  \n",
1790 |        "Software  Application  I applied through college or univ...  \n",
1791 |        "Software  Application  I applied through college or univ...  \n",
1792 |        "Software  Application  I applied through college or univ...  \n",
1793 |        "Software  Application  I applied online. The process too...  \n",
1794 |        "\n",
1795 |        "[17654 rows x 10 columns]"
1796 |       ]
1797 |      },
1798 |      "execution_count": 22,
1799 |      "metadata": {},
1800 |      "output_type": "execute_result"
1801 |     }
1802 |    ],
1803 |    "source": [
1804 |     "# Adding label \"Hardware\" if the company name is \"Orange\", \"Dell\", \"IDM\", or \"Siemens\".\n",
1805 |     "# \"Orange\" because we changed \"Apple\" to \"Orange\" above.\n",
1806 |     "# Adding label \"Software\" otherwise.\n",
1807 |     "\n",
1808 |     "def company_type(x):\n",
1809 |     "    hardware_companies = set([\"Orange\", \"Dell\", \"IBM\", \"Siemens\"])\n",
1810 |     "    return \"Hardware\" if x[\"Company\"] in hardware_companies else \"Software\"\n",
1811 |     "df[\"Type\"] = df.apply(lambda x: company_type(x), axis=1)\n",
1812 |     "\n",
1813 |     "# Setting \"Type\" to be labels. We call \"\"\n",
1814 |     "df = df.set_index(\"Type\")\n",
1815 |     "df\n",
1816 |     "# Label columns aren't considered part of the DataFrame's content.\n",
1817 |     "# After adding labels to your DataFrame, it still has 10 columns, same as before."
1818 |    ]
1819 |   },
1820 |   {
1821 |    "cell_type": "markdown",
1822 |    "metadata": {},
1823 |    "source": [
1824 |     "**Warning**: labels in `DataFrame` are stored as normal columns when you write the `DataFrame` to file using `.to_csv()`, and will need to be explicitly set after loading files, so if you send your CSV file to other people without explaining, they'll have no way of knowing which columns are labels. This might cause reproducibility issues. See [Stack Overflow answer](https://stackoverflow.com/questions/20109391/how-to-make-good-reproducible-pandas-examples)."
1825 |    ]
1826 |   },
1827 |   {
1828 |    "cell_type": "markdown",
1829 |    "metadata": {},
1830 |    "source": [
1831 |     "##### 3.2.2.1 Selecting rows by labels "
1832 |    ]
1833 |   },
1834 |   {
1835 |    "cell_type": "code",
1836 |    "execution_count": 23,
1837 |    "metadata": {},
1838 |    "outputs": [
1839 |     {
1840 |      "data": {
1841 |       "text/html": [
1842 |        "<div>\n",
1843 |        "<style scoped>\n",
1844 |        "    .dataframe tbody tr th:only-of-type {\n",
1845 |        "        vertical-align: middle;\n",
1846 |        "    }\n",
1847 |        "\n",
1848 |        "    .dataframe tbody tr th {\n",
1849 |        "        vertical-align: top;\n",
1850 |        "    }\n",
1851 |        "\n",
1852 |        "    .dataframe thead th {\n",
1853 |        "        text-align: right;\n",
1854 |        "    }\n",
1855 |        "</style>\n",
1856 |        "<table border=\"1\" class=\"dataframe\">\n",
1857 |        "  <thead>\n",
1858 |        "    <tr style=\"text-align: right;\">\n",
1859 |        "      <th></th>\n",
1860 |        "      <th>Company</th>\n",
1861 |        "      <th>Title</th>\n",
1862 |        "      <th>Job</th>\n",
1863 |        "      <th>Level</th>\n",
1864 |        "      <th>Date</th>\n",
1865 |        "      <th>Upvotes</th>\n",
1866 |        "      <th>Offer</th>\n",
1867 |        "      <th>Experience</th>\n",
1868 |        "      <th>Difficulty</th>\n",
1869 |        "      <th>Review</th>\n",
1870 |        "    </tr>\n",
1871 |        "    <tr>\n",
1872 |        "      <th>Type</th>\n",
1873 |        "      <th></th>\n",
1874 |        "      <th></th>\n",
1875 |        "      <th></th>\n",
1876 |        "      <th></th>\n",
1877 |        "      <th></th>\n",
1878 |        "      <th></th>\n",
1879 |        "      <th></th>\n",
1880 |        "      <th></th>\n",
1881 |        "      <th></th>\n",
1882 |        "      <th></th>\n",
1883 |        "    </tr>\n",
1884 |        "  </thead>\n",
1885 |        "  <tbody>\n",
1886 |        "    <tr>\n",
1887 |        "      <th>Hardware</th>\n",
1888 |        "      <td>Orange</td>\n",
1889 |        "      <td>Software Engineer</td>\n",
1890 |        "      <td>Software Engineer</td>\n",
1891 |        "      <td>Engineer</td>\n",
1892 |        "      <td>Aug 7, 2019</td>\n",
1893 |        "      <td>0</td>\n",
1894 |        "      <td>No offer</td>\n",
1895 |        "      <td>0.0</td>\n",
1896 |        "      <td>Medium</td>\n",
1897 |        "      <td>Orange is love. Orange is life.</td>\n",
1898 |        "    </tr>\n",
1899 |        "    <tr>\n",
1900 |        "      <th>Hardware</th>\n",
1901 |        "      <td>Orange</td>\n",
1902 |        "      <td>Software Engineer</td>\n",
1903 |        "      <td>Software Engineer</td>\n",
1904 |        "      <td>Engineer</td>\n",
1905 |        "      <td>Aug 8, 2019</td>\n",
1906 |        "      <td>0</td>\n",
1907 |        "      <td>Accepted offer</td>\n",
1908 |        "      <td>1.0</td>\n",
1909 |        "      <td>Hard</td>\n",
1910 |        "      <td>Application  I applied online. The process too...</td>\n",
1911 |        "    </tr>\n",
1912 |        "    <tr>\n",
1913 |        "      <th>Hardware</th>\n",
1914 |        "      <td>Orange</td>\n",
1915 |        "      <td>Software Engineer</td>\n",
1916 |        "      <td>Software Engineer</td>\n",
1917 |        "      <td>Engineer</td>\n",
1918 |        "      <td>NaN</td>\n",
1919 |        "      <td>0</td>\n",
1920 |        "      <td>Declined offer</td>\n",
1921 |        "      <td>0.0</td>\n",
1922 |        "      <td>Medium</td>\n",
1923 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
1924 |        "    </tr>\n",
1925 |        "    <tr>\n",
1926 |        "      <th>Hardware</th>\n",
1927 |        "      <td>Orange</td>\n",
1928 |        "      <td>Software Engineer</td>\n",
1929 |        "      <td>Software Engineer</td>\n",
1930 |        "      <td>Engineer</td>\n",
1931 |        "      <td>NaN</td>\n",
1932 |        "      <td>9</td>\n",
1933 |        "      <td>Declined offer</td>\n",
1934 |        "      <td>-1.0</td>\n",
1935 |        "      <td>Medium</td>\n",
1936 |        "      <td>Application  The process took a week. I interv...</td>\n",
1937 |        "    </tr>\n",
1938 |        "    <tr>\n",
1939 |        "      <th>Hardware</th>\n",
1940 |        "      <td>Orange</td>\n",
1941 |        "      <td>Software Engineer</td>\n",
1942 |        "      <td>Software Engineer</td>\n",
1943 |        "      <td>Engineer</td>\n",
1944 |        "      <td>May 29, 2009</td>\n",
1945 |        "      <td>2</td>\n",
1946 |        "      <td>No offer</td>\n",
1947 |        "      <td>0.0</td>\n",
1948 |        "      <td>Medium</td>\n",
1949 |        "      <td>Application  I applied through an employee ref...</td>\n",
1950 |        "    </tr>\n",
1951 |        "    <tr>\n",
1952 |        "      <th>...</th>\n",
1953 |        "      <td>...</td>\n",
1954 |        "      <td>...</td>\n",
1955 |        "      <td>...</td>\n",
1956 |        "      <td>...</td>\n",
1957 |        "      <td>...</td>\n",
1958 |        "      <td>...</td>\n",
1959 |        "      <td>...</td>\n",
1960 |        "      <td>...</td>\n",
1961 |        "      <td>...</td>\n",
1962 |        "      <td>...</td>\n",
1963 |        "    </tr>\n",
1964 |        "    <tr>\n",
1965 |        "      <th>Hardware</th>\n",
1966 |        "      <td>IBM</td>\n",
1967 |        "      <td>Senior Software Engineer</td>\n",
1968 |        "      <td>Software Engineer</td>\n",
1969 |        "      <td>Senior</td>\n",
1970 |        "      <td>Sep 20, 2015</td>\n",
1971 |        "      <td>0</td>\n",
1972 |        "      <td>No offer</td>\n",
1973 |        "      <td>-1.0</td>\n",
1974 |        "      <td>Easy</td>\n",
1975 |        "      <td>Application  I applied through a recruiter. Th...</td>\n",
1976 |        "    </tr>\n",
1977 |        "    <tr>\n",
1978 |        "      <th>Hardware</th>\n",
1979 |        "      <td>IBM</td>\n",
1980 |        "      <td>Senior Software Engineer</td>\n",
1981 |        "      <td>Software Engineer</td>\n",
1982 |        "      <td>Senior</td>\n",
1983 |        "      <td>Sep 14, 2015</td>\n",
1984 |        "      <td>0</td>\n",
1985 |        "      <td>Accepted offer</td>\n",
1986 |        "      <td>-1.0</td>\n",
1987 |        "      <td>Medium</td>\n",
1988 |        "      <td>Application  I applied in-person. The process ...</td>\n",
1989 |        "    </tr>\n",
1990 |        "    <tr>\n",
1991 |        "      <th>Hardware</th>\n",
1992 |        "      <td>IBM</td>\n",
1993 |        "      <td>Senior Software Engineer</td>\n",
1994 |        "      <td>Software Engineer</td>\n",
1995 |        "      <td>Senior</td>\n",
1996 |        "      <td>Aug 6, 2015</td>\n",
1997 |        "      <td>0</td>\n",
1998 |        "      <td>Accepted offer</td>\n",
1999 |        "      <td>1.0</td>\n",
2000 |        "      <td>Hard</td>\n",
2001 |        "      <td>Application  I applied through a recruiter. Th...</td>\n",
2002 |        "    </tr>\n",
2003 |        "    <tr>\n",
2004 |        "      <th>Hardware</th>\n",
2005 |        "      <td>IBM</td>\n",
2006 |        "      <td>Senior Software Engineer</td>\n",
2007 |        "      <td>Software Engineer</td>\n",
2008 |        "      <td>Senior</td>\n",
2009 |        "      <td>Dec 13, 2015</td>\n",
2010 |        "      <td>0</td>\n",
2011 |        "      <td>Accepted offer</td>\n",
2012 |        "      <td>1.0</td>\n",
2013 |        "      <td>Medium</td>\n",
2014 |        "      <td>Application  I applied online. The process too...</td>\n",
2015 |        "    </tr>\n",
2016 |        "    <tr>\n",
2017 |        "      <th>Hardware</th>\n",
2018 |        "      <td>IBM</td>\n",
2019 |        "      <td>Senior Software Engineer</td>\n",
2020 |        "      <td>Software Engineer</td>\n",
2021 |        "      <td>Senior</td>\n",
2022 |        "      <td>Feb 15, 2016</td>\n",
2023 |        "      <td>11</td>\n",
2024 |        "      <td>Accepted offer</td>\n",
2025 |        "      <td>1.0</td>\n",
2026 |        "      <td>Easy</td>\n",
2027 |        "      <td>Application  I applied online. The process too...</td>\n",
2028 |        "    </tr>\n",
2029 |        "  </tbody>\n",
2030 |        "</table>\n",
2031 |        "<p>1676 rows × 10 columns</p>\n",
2032 |        "</div>"
2033 |       ],
2034 |       "text/plain": [
2035 |        "         Company                     Title                Job     Level  \\\n",
2036 |        "Type                                                                      \n",
2037 |        "Hardware  Orange         Software Engineer  Software Engineer  Engineer   \n",
2038 |        "Hardware  Orange         Software Engineer  Software Engineer  Engineer   \n",
2039 |        "Hardware  Orange         Software Engineer  Software Engineer  Engineer   \n",
2040 |        "Hardware  Orange         Software Engineer  Software Engineer  Engineer   \n",
2041 |        "Hardware  Orange         Software Engineer  Software Engineer  Engineer   \n",
2042 |        "...          ...                       ...                ...       ...   \n",
2043 |        "Hardware     IBM  Senior Software Engineer  Software Engineer    Senior   \n",
2044 |        "Hardware     IBM  Senior Software Engineer  Software Engineer    Senior   \n",
2045 |        "Hardware     IBM  Senior Software Engineer  Software Engineer    Senior   \n",
2046 |        "Hardware     IBM  Senior Software Engineer  Software Engineer    Senior   \n",
2047 |        "Hardware     IBM  Senior Software Engineer  Software Engineer    Senior   \n",
2048 |        "\n",
2049 |        "                  Date  Upvotes           Offer  Experience Difficulty  \\\n",
2050 |        "Type                                                                     \n",
2051 |        "Hardware   Aug 7, 2019        0        No offer         0.0     Medium   \n",
2052 |        "Hardware   Aug 8, 2019        0  Accepted offer         1.0       Hard   \n",
2053 |        "Hardware           NaN        0  Declined offer         0.0     Medium   \n",
2054 |        "Hardware           NaN        9  Declined offer        -1.0     Medium   \n",
2055 |        "Hardware  May 29, 2009        2        No offer         0.0     Medium   \n",
2056 |        "...                ...      ...             ...         ...        ...   \n",
2057 |        "Hardware  Sep 20, 2015        0        No offer        -1.0       Easy   \n",
2058 |        "Hardware  Sep 14, 2015        0  Accepted offer        -1.0     Medium   \n",
2059 |        "Hardware   Aug 6, 2015        0  Accepted offer         1.0       Hard   \n",
2060 |        "Hardware  Dec 13, 2015        0  Accepted offer         1.0     Medium   \n",
2061 |        "Hardware  Feb 15, 2016       11  Accepted offer         1.0       Easy   \n",
2062 |        "\n",
2063 |        "                                                     Review  \n",
2064 |        "Type                                                         \n",
2065 |        "Hardware                    Orange is love. Orange is life.  \n",
2066 |        "Hardware  Application  I applied online. The process too...  \n",
2067 |        "Hardware  Application  The process took 4 weeks. I inter...  \n",
2068 |        "Hardware  Application  The process took a week. I interv...  \n",
2069 |        "Hardware  Application  I applied through an employee ref...  \n",
2070 |        "...                                                     ...  \n",
2071 |        "Hardware  Application  I applied through a recruiter. Th...  \n",
2072 |        "Hardware  Application  I applied in-person. The process ...  \n",
2073 |        "Hardware  Application  I applied through a recruiter. Th...  \n",
2074 |        "Hardware  Application  I applied online. The process too...  \n",
2075 |        "Hardware  Application  I applied online. The process too...  \n",
2076 |        "\n",
2077 |        "[1676 rows x 10 columns]"
2078 |       ]
2079 |      },
2080 |      "execution_count": 23,
2081 |      "metadata": {},
2082 |      "output_type": "execute_result"
2083 |     }
2084 |    ],
2085 |    "source": [
2086 |     "# Selecting rows with label \"Hardware\"\n",
2087 |     "df.loc[\"Hardware\"]"
2088 |    ]
2089 |   },
2090 |   {
2091 |    "cell_type": "code",
2092 |    "execution_count": 24,
2093 |    "metadata": {},
2094 |    "outputs": [
2095 |     {
2096 |      "data": {
2097 |       "text/html": [
2098 |        "<div>\n",
2099 |        "<style scoped>\n",
2100 |        "    .dataframe tbody tr th:only-of-type {\n",
2101 |        "        vertical-align: middle;\n",
2102 |        "    }\n",
2103 |        "\n",
2104 |        "    .dataframe tbody tr th {\n",
2105 |        "        vertical-align: top;\n",
2106 |        "    }\n",
2107 |        "\n",
2108 |        "    .dataframe thead th {\n",
2109 |        "        text-align: right;\n",
2110 |        "    }\n",
2111 |        "</style>\n",
2112 |        "<table border=\"1\" class=\"dataframe\">\n",
2113 |        "  <thead>\n",
2114 |        "    <tr style=\"text-align: right;\">\n",
2115 |        "      <th></th>\n",
2116 |        "      <th>Company</th>\n",
2117 |        "      <th>Title</th>\n",
2118 |        "      <th>Job</th>\n",
2119 |        "      <th>Level</th>\n",
2120 |        "      <th>Date</th>\n",
2121 |        "      <th>Upvotes</th>\n",
2122 |        "      <th>Offer</th>\n",
2123 |        "      <th>Experience</th>\n",
2124 |        "      <th>Difficulty</th>\n",
2125 |        "      <th>Review</th>\n",
2126 |        "    </tr>\n",
2127 |        "  </thead>\n",
2128 |        "  <tbody>\n",
2129 |        "    <tr>\n",
2130 |        "      <th>0</th>\n",
2131 |        "      <td>Orange</td>\n",
2132 |        "      <td>Software Engineer</td>\n",
2133 |        "      <td>Software Engineer</td>\n",
2134 |        "      <td>Engineer</td>\n",
2135 |        "      <td>Aug 7, 2019</td>\n",
2136 |        "      <td>0</td>\n",
2137 |        "      <td>No offer</td>\n",
2138 |        "      <td>0.0</td>\n",
2139 |        "      <td>Medium</td>\n",
2140 |        "      <td>Orange is love. Orange is life.</td>\n",
2141 |        "    </tr>\n",
2142 |        "    <tr>\n",
2143 |        "      <th>1</th>\n",
2144 |        "      <td>Orange</td>\n",
2145 |        "      <td>Software Engineer</td>\n",
2146 |        "      <td>Software Engineer</td>\n",
2147 |        "      <td>Engineer</td>\n",
2148 |        "      <td>Aug 8, 2019</td>\n",
2149 |        "      <td>0</td>\n",
2150 |        "      <td>Accepted offer</td>\n",
2151 |        "      <td>1.0</td>\n",
2152 |        "      <td>Hard</td>\n",
2153 |        "      <td>Application  I applied online. The process too...</td>\n",
2154 |        "    </tr>\n",
2155 |        "    <tr>\n",
2156 |        "      <th>2</th>\n",
2157 |        "      <td>Orange</td>\n",
2158 |        "      <td>Software Engineer</td>\n",
2159 |        "      <td>Software Engineer</td>\n",
2160 |        "      <td>Engineer</td>\n",
2161 |        "      <td>NaN</td>\n",
2162 |        "      <td>0</td>\n",
2163 |        "      <td>Declined offer</td>\n",
2164 |        "      <td>0.0</td>\n",
2165 |        "      <td>Medium</td>\n",
2166 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
2167 |        "    </tr>\n",
2168 |        "    <tr>\n",
2169 |        "      <th>3</th>\n",
2170 |        "      <td>Orange</td>\n",
2171 |        "      <td>Software Engineer</td>\n",
2172 |        "      <td>Software Engineer</td>\n",
2173 |        "      <td>Engineer</td>\n",
2174 |        "      <td>NaN</td>\n",
2175 |        "      <td>9</td>\n",
2176 |        "      <td>Declined offer</td>\n",
2177 |        "      <td>-1.0</td>\n",
2178 |        "      <td>Medium</td>\n",
2179 |        "      <td>Application  The process took a week. I interv...</td>\n",
2180 |        "    </tr>\n",
2181 |        "    <tr>\n",
2182 |        "      <th>4</th>\n",
2183 |        "      <td>Orange</td>\n",
2184 |        "      <td>Software Engineer</td>\n",
2185 |        "      <td>Software Engineer</td>\n",
2186 |        "      <td>Engineer</td>\n",
2187 |        "      <td>May 29, 2009</td>\n",
2188 |        "      <td>2</td>\n",
2189 |        "      <td>No offer</td>\n",
2190 |        "      <td>0.0</td>\n",
2191 |        "      <td>Medium</td>\n",
2192 |        "      <td>Application  I applied through an employee ref...</td>\n",
2193 |        "    </tr>\n",
2194 |        "    <tr>\n",
2195 |        "      <th>...</th>\n",
2196 |        "      <td>...</td>\n",
2197 |        "      <td>...</td>\n",
2198 |        "      <td>...</td>\n",
2199 |        "      <td>...</td>\n",
2200 |        "      <td>...</td>\n",
2201 |        "      <td>...</td>\n",
2202 |        "      <td>...</td>\n",
2203 |        "      <td>...</td>\n",
2204 |        "      <td>...</td>\n",
2205 |        "      <td>...</td>\n",
2206 |        "    </tr>\n",
2207 |        "    <tr>\n",
2208 |        "      <th>17649</th>\n",
2209 |        "      <td>Tencent</td>\n",
2210 |        "      <td>Software Engineer</td>\n",
2211 |        "      <td>Software Engineer</td>\n",
2212 |        "      <td>Engineer</td>\n",
2213 |        "      <td>May 25, 2012</td>\n",
2214 |        "      <td>0</td>\n",
2215 |        "      <td>Declined offer</td>\n",
2216 |        "      <td>0.0</td>\n",
2217 |        "      <td>Medium</td>\n",
2218 |        "      <td>Application  I applied online. The process too...</td>\n",
2219 |        "    </tr>\n",
2220 |        "    <tr>\n",
2221 |        "      <th>17650</th>\n",
2222 |        "      <td>Tencent</td>\n",
2223 |        "      <td>Software Engineer</td>\n",
2224 |        "      <td>Software Engineer</td>\n",
2225 |        "      <td>Engineer</td>\n",
2226 |        "      <td>Mar 15, 2014</td>\n",
2227 |        "      <td>0</td>\n",
2228 |        "      <td>No offer</td>\n",
2229 |        "      <td>NaN</td>\n",
2230 |        "      <td>NaN</td>\n",
2231 |        "      <td>Application  I applied through college or univ...</td>\n",
2232 |        "    </tr>\n",
2233 |        "    <tr>\n",
2234 |        "      <th>17651</th>\n",
2235 |        "      <td>Tencent</td>\n",
2236 |        "      <td>Software Engineer</td>\n",
2237 |        "      <td>Software Engineer</td>\n",
2238 |        "      <td>Engineer</td>\n",
2239 |        "      <td>Sep 22, 2015</td>\n",
2240 |        "      <td>0</td>\n",
2241 |        "      <td>Accepted offer</td>\n",
2242 |        "      <td>1.0</td>\n",
2243 |        "      <td>Medium</td>\n",
2244 |        "      <td>Application  I applied through college or univ...</td>\n",
2245 |        "    </tr>\n",
2246 |        "    <tr>\n",
2247 |        "      <th>17652</th>\n",
2248 |        "      <td>Tencent</td>\n",
2249 |        "      <td>Software Engineer</td>\n",
2250 |        "      <td>Software Engineer</td>\n",
2251 |        "      <td>Engineer</td>\n",
2252 |        "      <td>Jul 4, 2017</td>\n",
2253 |        "      <td>0</td>\n",
2254 |        "      <td>Declined offer</td>\n",
2255 |        "      <td>1.0</td>\n",
2256 |        "      <td>Medium</td>\n",
2257 |        "      <td>Application  I applied through college or univ...</td>\n",
2258 |        "    </tr>\n",
2259 |        "    <tr>\n",
2260 |        "      <th>17653</th>\n",
2261 |        "      <td>Tencent</td>\n",
2262 |        "      <td>Software Engineer</td>\n",
2263 |        "      <td>Software Engineer</td>\n",
2264 |        "      <td>Engineer</td>\n",
2265 |        "      <td>Sep 30, 2016</td>\n",
2266 |        "      <td>0</td>\n",
2267 |        "      <td>Declined offer</td>\n",
2268 |        "      <td>0.0</td>\n",
2269 |        "      <td>Easy</td>\n",
2270 |        "      <td>Application  I applied online. The process too...</td>\n",
2271 |        "    </tr>\n",
2272 |        "  </tbody>\n",
2273 |        "</table>\n",
2274 |        "<p>17654 rows × 10 columns</p>\n",
2275 |        "</div>"
2276 |       ],
2277 |       "text/plain": [
2278 |        "       Company              Title                Job     Level          Date  \\\n",
2279 |        "0       Orange  Software Engineer  Software Engineer  Engineer   Aug 7, 2019   \n",
2280 |        "1       Orange  Software Engineer  Software Engineer  Engineer   Aug 8, 2019   \n",
2281 |        "2       Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
2282 |        "3       Orange  Software Engineer  Software Engineer  Engineer           NaN   \n",
2283 |        "4       Orange  Software Engineer  Software Engineer  Engineer  May 29, 2009   \n",
2284 |        "...        ...                ...                ...       ...           ...   \n",
2285 |        "17649  Tencent  Software Engineer  Software Engineer  Engineer  May 25, 2012   \n",
2286 |        "17650  Tencent  Software Engineer  Software Engineer  Engineer  Mar 15, 2014   \n",
2287 |        "17651  Tencent  Software Engineer  Software Engineer  Engineer  Sep 22, 2015   \n",
2288 |        "17652  Tencent  Software Engineer  Software Engineer  Engineer   Jul 4, 2017   \n",
2289 |        "17653  Tencent  Software Engineer  Software Engineer  Engineer  Sep 30, 2016   \n",
2290 |        "\n",
2291 |        "       Upvotes           Offer  Experience Difficulty  \\\n",
2292 |        "0            0        No offer         0.0     Medium   \n",
2293 |        "1            0  Accepted offer         1.0       Hard   \n",
2294 |        "2            0  Declined offer         0.0     Medium   \n",
2295 |        "3            9  Declined offer        -1.0     Medium   \n",
2296 |        "4            2        No offer         0.0     Medium   \n",
2297 |        "...        ...             ...         ...        ...   \n",
2298 |        "17649        0  Declined offer         0.0     Medium   \n",
2299 |        "17650        0        No offer         NaN        NaN   \n",
2300 |        "17651        0  Accepted offer         1.0     Medium   \n",
2301 |        "17652        0  Declined offer         1.0     Medium   \n",
2302 |        "17653        0  Declined offer         0.0       Easy   \n",
2303 |        "\n",
2304 |        "                                                  Review  \n",
2305 |        "0                        Orange is love. Orange is life.  \n",
2306 |        "1      Application  I applied online. The process too...  \n",
2307 |        "2      Application  The process took 4 weeks. I inter...  \n",
2308 |        "3      Application  The process took a week. I interv...  \n",
2309 |        "4      Application  I applied through an employee ref...  \n",
2310 |        "...                                                  ...  \n",
2311 |        "17649  Application  I applied online. The process too...  \n",
2312 |        "17650  Application  I applied through college or univ...  \n",
2313 |        "17651  Application  I applied through college or univ...  \n",
2314 |        "17652  Application  I applied through college or univ...  \n",
2315 |        "17653  Application  I applied online. The process too...  \n",
2316 |        "\n",
2317 |        "[17654 rows x 10 columns]"
2318 |       ]
2319 |      },
2320 |      "execution_count": 24,
2321 |      "metadata": {},
2322 |      "output_type": "execute_result"
2323 |     }
2324 |    ],
2325 |    "source": [
2326 |     "# To drop a label, you need to use reset_index with drop=True\n",
2327 |     "df.reset_index(drop=True, inplace=True)\n",
2328 |     "df"
2329 |    ]
2330 |   },
2331 |   {
2332 |    "cell_type": "markdown",
2333 |    "metadata": {},
2334 |    "source": [
2335 |     "### 3.3 Slicing Series\n",
2336 |     "Slicing pandas `Series` is similar to slicing in Python."
2337 |    ]
2338 |   },
2339 |   {
2340 |    "cell_type": "code",
2341 |    "execution_count": 25,
2342 |    "metadata": {},
2343 |    "outputs": [
2344 |     {
2345 |      "data": {
2346 |       "text/plain": [
2347 |        "0      Orange\n",
2348 |        "100    Orange\n",
2349 |        "200    Orange\n",
2350 |        "300    Orange\n",
2351 |        "400     Intel\n",
2352 |        "500     Intel\n",
2353 |        "600     Intel\n",
2354 |        "700     Intel\n",
2355 |        "800      Uber\n",
2356 |        "900      Uber\n",
2357 |        "Name: Company, dtype: object"
2358 |       ]
2359 |      },
2360 |      "execution_count": 25,
2361 |      "metadata": {},
2362 |      "output_type": "execute_result"
2363 |     }
2364 |    ],
2365 |    "source": [
2366 |     "series = df.Company\n",
2367 |     "# The first 1000 companies, picking every 100th companies\n",
2368 |     "series[:1000:100]"
2369 |    ]
2370 |   },
2371 |   {
2372 |    "cell_type": "markdown",
2373 |    "metadata": {},
2374 |    "source": [
2375 |     "## 4. Accessors\n",
2376 |     "\n",
2377 |     "### 4.1 string accessor\n",
2378 |     "`.str` allows you to apply built-in string functions to all strings in a column (aka a pandas Series). These built-in functions come in handy when you want to do some basic string processing."
2379 |    ]
2380 |   },
2381 |   {
2382 |    "cell_type": "code",
2383 |    "execution_count": 26,
2384 |    "metadata": {},
2385 |    "outputs": [
2386 |     {
2387 |      "data": {
2388 |       "text/plain": [
2389 |        "0                          orange is love. orange is life.\n",
2390 |        "1        application  i applied online. the process too...\n",
2391 |        "2        application  the process took 4 weeks. i inter...\n",
2392 |        "3        application  the process took a week. i interv...\n",
2393 |        "4        application  i applied through an employee ref...\n",
2394 |        "                               ...                        \n",
2395 |        "17649    application  i applied online. the process too...\n",
2396 |        "17650    application  i applied through college or univ...\n",
2397 |        "17651    application  i applied through college or univ...\n",
2398 |        "17652    application  i applied through college or univ...\n",
2399 |        "17653    application  i applied online. the process too...\n",
2400 |        "Name: Review, Length: 17654, dtype: object"
2401 |       ]
2402 |      },
2403 |      "execution_count": 26,
2404 |      "metadata": {},
2405 |      "output_type": "execute_result"
2406 |     }
2407 |    ],
2408 |    "source": [
2409 |     "# If you want to lowercase all the reviews in the `Reviews` column.\n",
2410 |     "df[\"Review\"].str.lower()"
2411 |    ]
2412 |   },
2413 |   {
2414 |    "cell_type": "code",
2415 |    "execution_count": 27,
2416 |    "metadata": {},
2417 |    "outputs": [
2418 |     {
2419 |      "data": {
2420 |       "text/plain": [
2421 |        "0         31\n",
2422 |        "1        670\n",
2423 |        "2        350\n",
2424 |        "3        807\n",
2425 |        "4        663\n",
2426 |        "        ... \n",
2427 |        "17649    470\n",
2428 |        "17650    394\n",
2429 |        "17651    524\n",
2430 |        "17652    391\n",
2431 |        "17653    784\n",
2432 |        "Name: Review, Length: 17654, dtype: int64"
2433 |       ]
2434 |      },
2435 |      "execution_count": 27,
2436 |      "metadata": {},
2437 |      "output_type": "execute_result"
2438 |     }
2439 |    ],
2440 |    "source": [
2441 |     "# Or if you want to get the length of all the reviews\n",
2442 |     "df.Review.str.len()"
2443 |    ]
2444 |   },
2445 |   {
2446 |    "cell_type": "markdown",
2447 |    "metadata": {},
2448 |    "source": [
2449 |     "`.str` can be very powerful if you use it with Regex. Imagine you want to get a sense of how long the interview process takes for each review. You notice that each review mentions how long it takes such as \"the process took 4 weeks\". So you use this heuristic:\n",
2450 |     "- a process is short if it takes days\n",
2451 |     "- a process is average is if it takes weeks\n",
2452 |     "- a process is long if it takes at least 4 weeks"
2453 |    ]
2454 |   },
2455 |   {
2456 |    "cell_type": "code",
2457 |    "execution_count": 28,
2458 |    "metadata": {},
2459 |    "outputs": [
2460 |     {
2461 |      "data": {
2462 |       "text/html": [
2463 |        "<div>\n",
2464 |        "<style scoped>\n",
2465 |        "    .dataframe tbody tr th:only-of-type {\n",
2466 |        "        vertical-align: middle;\n",
2467 |        "    }\n",
2468 |        "\n",
2469 |        "    .dataframe tbody tr th {\n",
2470 |        "        vertical-align: top;\n",
2471 |        "    }\n",
2472 |        "\n",
2473 |        "    .dataframe thead th {\n",
2474 |        "        text-align: right;\n",
2475 |        "    }\n",
2476 |        "</style>\n",
2477 |        "<table border=\"1\" class=\"dataframe\">\n",
2478 |        "  <thead>\n",
2479 |        "    <tr style=\"text-align: right;\">\n",
2480 |        "      <th></th>\n",
2481 |        "      <th>Review</th>\n",
2482 |        "      <th>Process</th>\n",
2483 |        "    </tr>\n",
2484 |        "  </thead>\n",
2485 |        "  <tbody>\n",
2486 |        "    <tr>\n",
2487 |        "      <th>1</th>\n",
2488 |        "      <td>Application  I applied online. The process too...</td>\n",
2489 |        "      <td>Long</td>\n",
2490 |        "    </tr>\n",
2491 |        "    <tr>\n",
2492 |        "      <th>2</th>\n",
2493 |        "      <td>Application  The process took 4 weeks. I inter...</td>\n",
2494 |        "      <td>Long</td>\n",
2495 |        "    </tr>\n",
2496 |        "    <tr>\n",
2497 |        "      <th>3</th>\n",
2498 |        "      <td>Application  The process took a week. I interv...</td>\n",
2499 |        "      <td>Average</td>\n",
2500 |        "    </tr>\n",
2501 |        "    <tr>\n",
2502 |        "      <th>5</th>\n",
2503 |        "      <td>Application  I applied through college or univ...</td>\n",
2504 |        "      <td>Long</td>\n",
2505 |        "    </tr>\n",
2506 |        "    <tr>\n",
2507 |        "      <th>6</th>\n",
2508 |        "      <td>Application  The process took 2 days. I interv...</td>\n",
2509 |        "      <td>Short</td>\n",
2510 |        "    </tr>\n",
2511 |        "    <tr>\n",
2512 |        "      <th>...</th>\n",
2513 |        "      <td>...</td>\n",
2514 |        "      <td>...</td>\n",
2515 |        "    </tr>\n",
2516 |        "    <tr>\n",
2517 |        "      <th>17645</th>\n",
2518 |        "      <td>Application  I applied online. The process too...</td>\n",
2519 |        "      <td>Average</td>\n",
2520 |        "    </tr>\n",
2521 |        "    <tr>\n",
2522 |        "      <th>17647</th>\n",
2523 |        "      <td>Application  I applied through college or univ...</td>\n",
2524 |        "      <td>Average</td>\n",
2525 |        "    </tr>\n",
2526 |        "    <tr>\n",
2527 |        "      <th>17648</th>\n",
2528 |        "      <td>Application  I applied online. The process too...</td>\n",
2529 |        "      <td>Short</td>\n",
2530 |        "    </tr>\n",
2531 |        "    <tr>\n",
2532 |        "      <th>17649</th>\n",
2533 |        "      <td>Application  I applied online. The process too...</td>\n",
2534 |        "      <td>Average</td>\n",
2535 |        "    </tr>\n",
2536 |        "    <tr>\n",
2537 |        "      <th>17653</th>\n",
2538 |        "      <td>Application  I applied online. The process too...</td>\n",
2539 |        "      <td>Average</td>\n",
2540 |        "    </tr>\n",
2541 |        "  </tbody>\n",
2542 |        "</table>\n",
2543 |        "<p>12045 rows × 2 columns</p>\n",
2544 |        "</div>"
2545 |       ],
2546 |       "text/plain": [
2547 |        "                                                  Review  Process\n",
2548 |        "1      Application  I applied online. The process too...     Long\n",
2549 |        "2      Application  The process took 4 weeks. I inter...     Long\n",
2550 |        "3      Application  The process took a week. I interv...  Average\n",
2551 |        "5      Application  I applied through college or univ...     Long\n",
2552 |        "6      Application  The process took 2 days. I interv...    Short\n",
2553 |        "...                                                  ...      ...\n",
2554 |        "17645  Application  I applied online. The process too...  Average\n",
2555 |        "17647  Application  I applied through college or univ...  Average\n",
2556 |        "17648  Application  I applied online. The process too...    Short\n",
2557 |        "17649  Application  I applied online. The process too...  Average\n",
2558 |        "17653  Application  I applied online. The process too...  Average\n",
2559 |        "\n",
2560 |        "[12045 rows x 2 columns]"
2561 |       ]
2562 |      },
2563 |      "execution_count": 28,
2564 |      "metadata": {},
2565 |      "output_type": "execute_result"
2566 |     }
2567 |    ],
2568 |    "source": [
2569 |     "df.loc[df[\"Review\"].str.contains(\"days\"), \"Process\"] = \"Short\"\n",
2570 |     "df.loc[df[\"Review\"].str.contains(\"week\"), \"Process\"] = \"Average\"\n",
2571 |     "df.loc[df[\"Review\"].str.contains(\"month|[4-9]+[^ ]* weeks|[1-9]\\d{1,}[^ ]* weeks\"), \"Process\"] = \"Long\"\n",
2572 |     "df[~df.Process.isna()][[\"Review\", \"Process\"]]"
2573 |    ]
2574 |   },
2575 |   {
2576 |    "cell_type": "markdown",
2577 |    "metadata": {},
2578 |    "source": [
2579 |     "We want to sanity check if `Process` corresponds to `Review`, but `Review` is cut off in the display above. To show wider columns, you can set `display.max_colwidth` to `100`.\n",
2580 |     "\n",
2581 |     "**Note**: set_option has several great options you should check out."
2582 |    ]
2583 |   },
2584 |   {
2585 |    "cell_type": "code",
2586 |    "execution_count": 29,
2587 |    "metadata": {},
2588 |    "outputs": [
2589 |     {
2590 |      "data": {
2591 |       "text/html": [
2592 |        "<div>\n",
2593 |        "<style scoped>\n",
2594 |        "    .dataframe tbody tr th:only-of-type {\n",
2595 |        "        vertical-align: middle;\n",
2596 |        "    }\n",
2597 |        "\n",
2598 |        "    .dataframe tbody tr th {\n",
2599 |        "        vertical-align: top;\n",
2600 |        "    }\n",
2601 |        "\n",
2602 |        "    .dataframe thead th {\n",
2603 |        "        text-align: right;\n",
2604 |        "    }\n",
2605 |        "</style>\n",
2606 |        "<table border=\"1\" class=\"dataframe\">\n",
2607 |        "  <thead>\n",
2608 |        "    <tr style=\"text-align: right;\">\n",
2609 |        "      <th></th>\n",
2610 |        "      <th>Review</th>\n",
2611 |        "      <th>Process</th>\n",
2612 |        "    </tr>\n",
2613 |        "  </thead>\n",
2614 |        "  <tbody>\n",
2615 |        "    <tr>\n",
2616 |        "      <th>1</th>\n",
2617 |        "      <td>Application  I applied online. The process took 2+ months. I interviewed at Apple (San Jose, CA)...</td>\n",
2618 |        "      <td>Long</td>\n",
2619 |        "    </tr>\n",
2620 |        "    <tr>\n",
2621 |        "      <th>2</th>\n",
2622 |        "      <td>Application  The process took 4 weeks. I interviewed at Apple (San Antonio, TX) in February 2016...</td>\n",
2623 |        "      <td>Long</td>\n",
2624 |        "    </tr>\n",
2625 |        "    <tr>\n",
2626 |        "      <th>3</th>\n",
2627 |        "      <td>Application  The process took a week. I interviewed at Apple (Cupertino, CA) in December 2008.  ...</td>\n",
2628 |        "      <td>Average</td>\n",
2629 |        "    </tr>\n",
2630 |        "    <tr>\n",
2631 |        "      <th>5</th>\n",
2632 |        "      <td>Application  I applied through college or university. The process took 6 weeks. I interviewed at...</td>\n",
2633 |        "      <td>Long</td>\n",
2634 |        "    </tr>\n",
2635 |        "    <tr>\n",
2636 |        "      <th>6</th>\n",
2637 |        "      <td>Application  The process took 2 days. I interviewed at Apple (Cupertino, CA) in March 2009.  Int...</td>\n",
2638 |        "      <td>Short</td>\n",
2639 |        "    </tr>\n",
2640 |        "    <tr>\n",
2641 |        "      <th>...</th>\n",
2642 |        "      <td>...</td>\n",
2643 |        "      <td>...</td>\n",
2644 |        "    </tr>\n",
2645 |        "    <tr>\n",
2646 |        "      <th>17645</th>\n",
2647 |        "      <td>Application  I applied online. The process took a week. I interviewed at Tencent (Palo Alto, CA)...</td>\n",
2648 |        "      <td>Average</td>\n",
2649 |        "    </tr>\n",
2650 |        "    <tr>\n",
2651 |        "      <th>17647</th>\n",
2652 |        "      <td>Application  I applied through college or university. The process took 4+ weeks. I interviewed a...</td>\n",
2653 |        "      <td>Average</td>\n",
2654 |        "    </tr>\n",
2655 |        "    <tr>\n",
2656 |        "      <th>17648</th>\n",
2657 |        "      <td>Application  I applied online. The process took 2 days. I interviewed at Tencent.  Interview  I ...</td>\n",
2658 |        "      <td>Short</td>\n",
2659 |        "    </tr>\n",
2660 |        "    <tr>\n",
2661 |        "      <th>17649</th>\n",
2662 |        "      <td>Application  I applied online. The process took a week. I interviewed at Tencent (Beijing, Beiji...</td>\n",
2663 |        "      <td>Average</td>\n",
2664 |        "    </tr>\n",
2665 |        "    <tr>\n",
2666 |        "      <th>17653</th>\n",
2667 |        "      <td>Application  I applied online. The process took 3+ weeks. I interviewed at Tencent (Beijing, Bei...</td>\n",
2668 |        "      <td>Average</td>\n",
2669 |        "    </tr>\n",
2670 |        "  </tbody>\n",
2671 |        "</table>\n",
2672 |        "<p>12045 rows × 2 columns</p>\n",
2673 |        "</div>"
2674 |       ],
2675 |       "text/plain": [
2676 |        "                                                                                                    Review  \\\n",
2677 |        "1      Application  I applied online. The process took 2+ months. I interviewed at Apple (San Jose, CA)...   \n",
2678 |        "2      Application  The process took 4 weeks. I interviewed at Apple (San Antonio, TX) in February 2016...   \n",
2679 |        "3      Application  The process took a week. I interviewed at Apple (Cupertino, CA) in December 2008.  ...   \n",
2680 |        "5      Application  I applied through college or university. The process took 6 weeks. I interviewed at...   \n",
2681 |        "6      Application  The process took 2 days. I interviewed at Apple (Cupertino, CA) in March 2009.  Int...   \n",
2682 |        "...                                                                                                    ...   \n",
2683 |        "17645  Application  I applied online. The process took a week. I interviewed at Tencent (Palo Alto, CA)...   \n",
2684 |        "17647  Application  I applied through college or university. The process took 4+ weeks. I interviewed a...   \n",
2685 |        "17648  Application  I applied online. The process took 2 days. I interviewed at Tencent.  Interview  I ...   \n",
2686 |        "17649  Application  I applied online. The process took a week. I interviewed at Tencent (Beijing, Beiji...   \n",
2687 |        "17653  Application  I applied online. The process took 3+ weeks. I interviewed at Tencent (Beijing, Bei...   \n",
2688 |        "\n",
2689 |        "       Process  \n",
2690 |        "1         Long  \n",
2691 |        "2         Long  \n",
2692 |        "3      Average  \n",
2693 |        "5         Long  \n",
2694 |        "6        Short  \n",
2695 |        "...        ...  \n",
2696 |        "17645  Average  \n",
2697 |        "17647  Average  \n",
2698 |        "17648    Short  \n",
2699 |        "17649  Average  \n",
2700 |        "17653  Average  \n",
2701 |        "\n",
2702 |        "[12045 rows x 2 columns]"
2703 |       ]
2704 |      },
2705 |      "execution_count": 29,
2706 |      "metadata": {},
2707 |      "output_type": "execute_result"
2708 |     }
2709 |    ],
2710 |    "source": [
2711 |     "pd.set_option('display.max_colwidth', 100)\n",
2712 |     "df[~df.Process.isna()][[\"Review\", \"Process\"]]"
2713 |    ]
2714 |   },
2715 |   {
2716 |    "cell_type": "code",
2717 |    "execution_count": 30,
2718 |    "metadata": {},
2719 |    "outputs": [
2720 |     {
2721 |      "data": {
2722 |       "text/plain": [
2723 |        "dict_keys(['__module__', '__annotations__', '__doc__', '__init__', '_validate', '__getitem__', '__iter__', '_wrap_result', '_get_series_list', 'cat', 'split', 'rsplit', 'partition', 'rpartition', 'get', 'join', 'contains', 'match', 'fullmatch', 'replace', 'repeat', 'pad', 'center', 'ljust', 'rjust', 'zfill', 'slice', 'slice_replace', 'decode', 'encode', 'strip', 'lstrip', 'rstrip', 'wrap', 'get_dummies', 'translate', 'count', 'startswith', 'endswith', 'findall', 'extract', 'extractall', 'find', 'rfind', 'normalize', 'index', 'rindex', 'len', '_doc_args', 'lower', 'upper', 'title', 'capitalize', 'swapcase', 'casefold', 'isalnum', 'isalpha', 'isdigit', 'isspace', 'islower', 'isupper', 'istitle', 'isnumeric', 'isdecimal', '_make_accessor'])"
2724 |       ]
2725 |      },
2726 |      "execution_count": 30,
2727 |      "metadata": {},
2728 |      "output_type": "execute_result"
2729 |     }
2730 |    ],
2731 |    "source": [
2732 |     "# To see the built-in functions available for `.str`, use this\n",
2733 |     "pd.Series.str.__dict__.keys()"
2734 |    ]
2735 |   },
2736 |   {
2737 |    "cell_type": "markdown",
2738 |    "metadata": {},
2739 |    "source": [
2740 |     "### 4.2 Other accessors\n",
2741 |     "pandas `Series` has 3 other accessors.\n",
2742 |     "- `.dt`: handles date formats\n",
2743 |     "- `.cat`: handles categorical data\n",
2744 |     "- `.sparse`: handles sparse matrices"
2745 |    ]
2746 |   },
2747 |   {
2748 |    "cell_type": "code",
2749 |    "execution_count": 31,
2750 |    "metadata": {},
2751 |    "outputs": [
2752 |     {
2753 |      "data": {
2754 |       "text/plain": [
2755 |        "{'cat', 'dt', 'sparse', 'str'}"
2756 |       ]
2757 |      },
2758 |      "execution_count": 31,
2759 |      "metadata": {},
2760 |      "output_type": "execute_result"
2761 |     }
2762 |    ],
2763 |    "source": [
2764 |     "pd.Series._accessors"
2765 |    ]
2766 |   },
2767 |   {
2768 |    "cell_type": "markdown",
2769 |    "metadata": {},
2770 |    "source": [
2771 |     "## 5. Data exploration\n",
2772 |     "When analyzing data, you might want to take a look at the data. pandas has some great built-in functions for that.\n",
2773 |     "\n",
2774 |     "### 5.1 `.head()`, `.tail()`, `.describe()`, `.info()`\n",
2775 |     "You're probably familiar with `.head()` and `.tail()` methods for showing the first/last rows of `DataFrame`. By default, 5 rows are shown, but you can specify the exact number."
2776 |    ]
2777 |   },
2778 |   {
2779 |    "cell_type": "code",
2780 |    "execution_count": 32,
2781 |    "metadata": {},
2782 |    "outputs": [
2783 |     {
2784 |      "data": {
2785 |       "text/html": [
2786 |        "<div>\n",
2787 |        "<style scoped>\n",
2788 |        "    .dataframe tbody tr th:only-of-type {\n",
2789 |        "        vertical-align: middle;\n",
2790 |        "    }\n",
2791 |        "\n",
2792 |        "    .dataframe tbody tr th {\n",
2793 |        "        vertical-align: top;\n",
2794 |        "    }\n",
2795 |        "\n",
2796 |        "    .dataframe thead th {\n",
2797 |        "        text-align: right;\n",
2798 |        "    }\n",
2799 |        "</style>\n",
2800 |        "<table border=\"1\" class=\"dataframe\">\n",
2801 |        "  <thead>\n",
2802 |        "    <tr style=\"text-align: right;\">\n",
2803 |        "      <th></th>\n",
2804 |        "      <th>Company</th>\n",
2805 |        "      <th>Title</th>\n",
2806 |        "      <th>Job</th>\n",
2807 |        "      <th>Level</th>\n",
2808 |        "      <th>Date</th>\n",
2809 |        "      <th>Upvotes</th>\n",
2810 |        "      <th>Offer</th>\n",
2811 |        "      <th>Experience</th>\n",
2812 |        "      <th>Difficulty</th>\n",
2813 |        "      <th>Review</th>\n",
2814 |        "      <th>Process</th>\n",
2815 |        "    </tr>\n",
2816 |        "  </thead>\n",
2817 |        "  <tbody>\n",
2818 |        "    <tr>\n",
2819 |        "      <th>17646</th>\n",
2820 |        "      <td>Tencent</td>\n",
2821 |        "      <td>Software Engineer</td>\n",
2822 |        "      <td>Software Engineer</td>\n",
2823 |        "      <td>Engineer</td>\n",
2824 |        "      <td>May 28, 2010</td>\n",
2825 |        "      <td>0</td>\n",
2826 |        "      <td>Declined offer</td>\n",
2827 |        "      <td>0.0</td>\n",
2828 |        "      <td>Easy</td>\n",
2829 |        "      <td>Application  I applied through an employee referral. The process took 1 day. I interviewed at Te...</td>\n",
2830 |        "      <td>NaN</td>\n",
2831 |        "    </tr>\n",
2832 |        "    <tr>\n",
2833 |        "      <th>17647</th>\n",
2834 |        "      <td>Tencent</td>\n",
2835 |        "      <td>Software Engineer</td>\n",
2836 |        "      <td>Software Engineer</td>\n",
2837 |        "      <td>Engineer</td>\n",
2838 |        "      <td>Apr 11, 2019</td>\n",
2839 |        "      <td>0</td>\n",
2840 |        "      <td>Accepted offer</td>\n",
2841 |        "      <td>1.0</td>\n",
2842 |        "      <td>Medium</td>\n",
2843 |        "      <td>Application  I applied through college or university. The process took 4+ weeks. I interviewed a...</td>\n",
2844 |        "      <td>Average</td>\n",
2845 |        "    </tr>\n",
2846 |        "    <tr>\n",
2847 |        "      <th>17648</th>\n",
2848 |        "      <td>Tencent</td>\n",
2849 |        "      <td>Software Engineer</td>\n",
2850 |        "      <td>Software Engineer</td>\n",
2851 |        "      <td>Engineer</td>\n",
2852 |        "      <td>Nov 4, 2012</td>\n",
2853 |        "      <td>0</td>\n",
2854 |        "      <td>No offer</td>\n",
2855 |        "      <td>NaN</td>\n",
2856 |        "      <td>NaN</td>\n",
2857 |        "      <td>Application  I applied online. The process took 2 days. I interviewed at Tencent.  Interview  I ...</td>\n",
2858 |        "      <td>Short</td>\n",
2859 |        "    </tr>\n",
2860 |        "    <tr>\n",
2861 |        "      <th>17649</th>\n",
2862 |        "      <td>Tencent</td>\n",
2863 |        "      <td>Software Engineer</td>\n",
2864 |        "      <td>Software Engineer</td>\n",
2865 |        "      <td>Engineer</td>\n",
2866 |        "      <td>May 25, 2012</td>\n",
2867 |        "      <td>0</td>\n",
2868 |        "      <td>Declined offer</td>\n",
2869 |        "      <td>0.0</td>\n",
2870 |        "      <td>Medium</td>\n",
2871 |        "      <td>Application  I applied online. The process took a week. I interviewed at Tencent (Beijing, Beiji...</td>\n",
2872 |        "      <td>Average</td>\n",
2873 |        "    </tr>\n",
2874 |        "    <tr>\n",
2875 |        "      <th>17650</th>\n",
2876 |        "      <td>Tencent</td>\n",
2877 |        "      <td>Software Engineer</td>\n",
2878 |        "      <td>Software Engineer</td>\n",
2879 |        "      <td>Engineer</td>\n",
2880 |        "      <td>Mar 15, 2014</td>\n",
2881 |        "      <td>0</td>\n",
2882 |        "      <td>No offer</td>\n",
2883 |        "      <td>NaN</td>\n",
2884 |        "      <td>NaN</td>\n",
2885 |        "      <td>Application  I applied through college or university. I interviewed at Tencent.  Interview  Prof...</td>\n",
2886 |        "      <td>NaN</td>\n",
2887 |        "    </tr>\n",
2888 |        "    <tr>\n",
2889 |        "      <th>17651</th>\n",
2890 |        "      <td>Tencent</td>\n",
2891 |        "      <td>Software Engineer</td>\n",
2892 |        "      <td>Software Engineer</td>\n",
2893 |        "      <td>Engineer</td>\n",
2894 |        "      <td>Sep 22, 2015</td>\n",
2895 |        "      <td>0</td>\n",
2896 |        "      <td>Accepted offer</td>\n",
2897 |        "      <td>1.0</td>\n",
2898 |        "      <td>Medium</td>\n",
2899 |        "      <td>Application  I applied through college or university. The process took 1 day. I interviewed at T...</td>\n",
2900 |        "      <td>NaN</td>\n",
2901 |        "    </tr>\n",
2902 |        "    <tr>\n",
2903 |        "      <th>17652</th>\n",
2904 |        "      <td>Tencent</td>\n",
2905 |        "      <td>Software Engineer</td>\n",
2906 |        "      <td>Software Engineer</td>\n",
2907 |        "      <td>Engineer</td>\n",
2908 |        "      <td>Jul 4, 2017</td>\n",
2909 |        "      <td>0</td>\n",
2910 |        "      <td>Declined offer</td>\n",
2911 |        "      <td>1.0</td>\n",
2912 |        "      <td>Medium</td>\n",
2913 |        "      <td>Application  I applied through college or university. I interviewed at Tencent (London, England ...</td>\n",
2914 |        "      <td>NaN</td>\n",
2915 |        "    </tr>\n",
2916 |        "    <tr>\n",
2917 |        "      <th>17653</th>\n",
2918 |        "      <td>Tencent</td>\n",
2919 |        "      <td>Software Engineer</td>\n",
2920 |        "      <td>Software Engineer</td>\n",
2921 |        "      <td>Engineer</td>\n",
2922 |        "      <td>Sep 30, 2016</td>\n",
2923 |        "      <td>0</td>\n",
2924 |        "      <td>Declined offer</td>\n",
2925 |        "      <td>0.0</td>\n",
2926 |        "      <td>Easy</td>\n",
2927 |        "      <td>Application  I applied online. The process took 3+ weeks. I interviewed at Tencent (Beijing, Bei...</td>\n",
2928 |        "      <td>Average</td>\n",
2929 |        "    </tr>\n",
2930 |        "  </tbody>\n",
2931 |        "</table>\n",
2932 |        "</div>"
2933 |       ],
2934 |       "text/plain": [
2935 |        "       Company              Title                Job     Level          Date  \\\n",
2936 |        "17646  Tencent  Software Engineer  Software Engineer  Engineer  May 28, 2010   \n",
2937 |        "17647  Tencent  Software Engineer  Software Engineer  Engineer  Apr 11, 2019   \n",
2938 |        "17648  Tencent  Software Engineer  Software Engineer  Engineer   Nov 4, 2012   \n",
2939 |        "17649  Tencent  Software Engineer  Software Engineer  Engineer  May 25, 2012   \n",
2940 |        "17650  Tencent  Software Engineer  Software Engineer  Engineer  Mar 15, 2014   \n",
2941 |        "17651  Tencent  Software Engineer  Software Engineer  Engineer  Sep 22, 2015   \n",
2942 |        "17652  Tencent  Software Engineer  Software Engineer  Engineer   Jul 4, 2017   \n",
2943 |        "17653  Tencent  Software Engineer  Software Engineer  Engineer  Sep 30, 2016   \n",
2944 |        "\n",
2945 |        "       Upvotes           Offer  Experience Difficulty  \\\n",
2946 |        "17646        0  Declined offer         0.0       Easy   \n",
2947 |        "17647        0  Accepted offer         1.0     Medium   \n",
2948 |        "17648        0        No offer         NaN        NaN   \n",
2949 |        "17649        0  Declined offer         0.0     Medium   \n",
2950 |        "17650        0        No offer         NaN        NaN   \n",
2951 |        "17651        0  Accepted offer         1.0     Medium   \n",
2952 |        "17652        0  Declined offer         1.0     Medium   \n",
2953 |        "17653        0  Declined offer         0.0       Easy   \n",
2954 |        "\n",
2955 |        "                                                                                                    Review  \\\n",
2956 |        "17646  Application  I applied through an employee referral. The process took 1 day. I interviewed at Te...   \n",
2957 |        "17647  Application  I applied through college or university. The process took 4+ weeks. I interviewed a...   \n",
2958 |        "17648  Application  I applied online. The process took 2 days. I interviewed at Tencent.  Interview  I ...   \n",
2959 |        "17649  Application  I applied online. The process took a week. I interviewed at Tencent (Beijing, Beiji...   \n",
2960 |        "17650  Application  I applied through college or university. I interviewed at Tencent.  Interview  Prof...   \n",
2961 |        "17651  Application  I applied through college or university. The process took 1 day. I interviewed at T...   \n",
2962 |        "17652  Application  I applied through college or university. I interviewed at Tencent (London, England ...   \n",
2963 |        "17653  Application  I applied online. The process took 3+ weeks. I interviewed at Tencent (Beijing, Bei...   \n",
2964 |        "\n",
2965 |        "       Process  \n",
2966 |        "17646      NaN  \n",
2967 |        "17647  Average  \n",
2968 |        "17648    Short  \n",
2969 |        "17649  Average  \n",
2970 |        "17650      NaN  \n",
2971 |        "17651      NaN  \n",
2972 |        "17652      NaN  \n",
2973 |        "17653  Average  "
2974 |       ]
2975 |      },
2976 |      "execution_count": 32,
2977 |      "metadata": {},
2978 |      "output_type": "execute_result"
2979 |     }
2980 |    ],
2981 |    "source": [
2982 |     "df.tail(8)"
2983 |    ]
2984 |   },
2985 |   {
2986 |    "cell_type": "code",
2987 |    "execution_count": 33,
2988 |    "metadata": {},
2989 |    "outputs": [
2990 |     {
2991 |      "data": {
2992 |       "text/html": [
2993 |        "<div>\n",
2994 |        "<style scoped>\n",
2995 |        "    .dataframe tbody tr th:only-of-type {\n",
2996 |        "        vertical-align: middle;\n",
2997 |        "    }\n",
2998 |        "\n",
2999 |        "    .dataframe tbody tr th {\n",
3000 |        "        vertical-align: top;\n",
3001 |        "    }\n",
3002 |        "\n",
3003 |        "    .dataframe thead th {\n",
3004 |        "        text-align: right;\n",
3005 |        "    }\n",
3006 |        "</style>\n",
3007 |        "<table border=\"1\" class=\"dataframe\">\n",
3008 |        "  <thead>\n",
3009 |        "    <tr style=\"text-align: right;\">\n",
3010 |        "      <th></th>\n",
3011 |        "      <th>Upvotes</th>\n",
3012 |        "      <th>Experience</th>\n",
3013 |        "    </tr>\n",
3014 |        "  </thead>\n",
3015 |        "  <tbody>\n",
3016 |        "    <tr>\n",
3017 |        "      <th>count</th>\n",
3018 |        "      <td>17654.000000</td>\n",
3019 |        "      <td>16365.000000</td>\n",
3020 |        "    </tr>\n",
3021 |        "    <tr>\n",
3022 |        "      <th>mean</th>\n",
3023 |        "      <td>2.298459</td>\n",
3024 |        "      <td>0.431714</td>\n",
3025 |        "    </tr>\n",
3026 |        "    <tr>\n",
3027 |        "      <th>std</th>\n",
3028 |        "      <td>28.252562</td>\n",
3029 |        "      <td>0.759964</td>\n",
3030 |        "    </tr>\n",
3031 |        "    <tr>\n",
3032 |        "      <th>min</th>\n",
3033 |        "      <td>0.000000</td>\n",
3034 |        "      <td>-1.000000</td>\n",
3035 |        "    </tr>\n",
3036 |        "    <tr>\n",
3037 |        "      <th>25%</th>\n",
3038 |        "      <td>0.000000</td>\n",
3039 |        "      <td>0.000000</td>\n",
3040 |        "    </tr>\n",
3041 |        "    <tr>\n",
3042 |        "      <th>50%</th>\n",
3043 |        "      <td>0.000000</td>\n",
3044 |        "      <td>1.000000</td>\n",
3045 |        "    </tr>\n",
3046 |        "    <tr>\n",
3047 |        "      <th>75%</th>\n",
3048 |        "      <td>1.000000</td>\n",
3049 |        "      <td>1.000000</td>\n",
3050 |        "    </tr>\n",
3051 |        "    <tr>\n",
3052 |        "      <th>max</th>\n",
3053 |        "      <td>1916.000000</td>\n",
3054 |        "      <td>1.000000</td>\n",
3055 |        "    </tr>\n",
3056 |        "  </tbody>\n",
3057 |        "</table>\n",
3058 |        "</div>"
3059 |       ],
3060 |       "text/plain": [
3061 |        "            Upvotes    Experience\n",
3062 |        "count  17654.000000  16365.000000\n",
3063 |        "mean       2.298459      0.431714\n",
3064 |        "std       28.252562      0.759964\n",
3065 |        "min        0.000000     -1.000000\n",
3066 |        "25%        0.000000      0.000000\n",
3067 |        "50%        0.000000      1.000000\n",
3068 |        "75%        1.000000      1.000000\n",
3069 |        "max     1916.000000      1.000000"
3070 |       ]
3071 |      },
3072 |      "execution_count": 33,
3073 |      "metadata": {},
3074 |      "output_type": "execute_result"
3075 |     }
3076 |    ],
3077 |    "source": [
3078 |     "# Generate statistics about numeric columns.\n",
3079 |     "df.describe()\n",
3080 |     "\n",
3081 |     "# Note:\n",
3082 |     "# 1. `.describe()` ignores all non-numeric columns.\n",
3083 |     "# 2. It doesn't take into account NaN values. So, the number shown in `count` below is the number of non-NaN entries."
3084 |    ]
3085 |   },
3086 |   {
3087 |    "cell_type": "code",
3088 |    "execution_count": 34,
3089 |    "metadata": {},
3090 |    "outputs": [
3091 |     {
3092 |      "name": "stdout",
3093 |      "output_type": "stream",
3094 |      "text": [
3095 |       "<class 'pandas.core.frame.DataFrame'>\n",
3096 |       "RangeIndex: 17654 entries, 0 to 17653\n",
3097 |       "Data columns (total 11 columns):\n",
3098 |       " #   Column      Non-Null Count  Dtype  \n",
3099 |       "---  ------      --------------  -----  \n",
3100 |       " 0   Company     17654 non-null  object \n",
3101 |       " 1   Title       17654 non-null  object \n",
3102 |       " 2   Job         17654 non-null  object \n",
3103 |       " 3   Level       17654 non-null  object \n",
3104 |       " 4   Date        17652 non-null  object \n",
3105 |       " 5   Upvotes     17654 non-null  int64  \n",
3106 |       " 6   Offer       17654 non-null  object \n",
3107 |       " 7   Experience  16365 non-null  float64\n",
3108 |       " 8   Difficulty  16376 non-null  object \n",
3109 |       " 9   Review      17654 non-null  object \n",
3110 |       " 10  Process     12045 non-null  object \n",
3111 |       "dtypes: float64(1), int64(1), object(9)\n",
3112 |       "memory usage: 1.5+ MB\n"
3113 |      ]
3114 |     }
3115 |    ],
3116 |    "source": [
3117 |     "# Show non-null count and types of all columns\n",
3118 |     "df.info()\n",
3119 |     "\n",
3120 |     "# Note: pandas treats string type as object"
3121 |    ]
3122 |   },
3123 |   {
3124 |    "cell_type": "code",
3125 |    "execution_count": 35,
3126 |    "metadata": {},
3127 |    "outputs": [
3128 |     {
3129 |      "data": {
3130 |       "text/plain": [
3131 |        "Company        1126274\n",
3132 |        "Title          1358305\n",
3133 |        "Job            1345085\n",
3134 |        "Level          1144784\n",
3135 |        "Date           1212964\n",
3136 |        "Upvotes         141384\n",
3137 |        "Offer          1184190\n",
3138 |        "Experience      141384\n",
3139 |        "Difficulty     1058008\n",
3140 |        "Review        20503467\n",
3141 |        "Process        1274180\n",
3142 |        "dtype: int64"
3143 |       ]
3144 |      },
3145 |      "execution_count": 35,
3146 |      "metadata": {},
3147 |      "output_type": "execute_result"
3148 |     }
3149 |    ],
3150 |    "source": [
3151 |     "# You can also see how much space your DataFrame is taking up\n",
3152 |     "import sys\n",
3153 |     "df.apply(sys.getsizeof)"
3154 |    ]
3155 |   },
3156 |   {
3157 |    "cell_type": "markdown",
3158 |    "metadata": {},
3159 |    "source": [
3160 |     "### 5.2 Count unique values\n",
3161 |     "You can get the number of unique values in a row (excluding NaN) with `nunique()`."
3162 |    ]
3163 |   },
3164 |   {
3165 |    "cell_type": "code",
3166 |    "execution_count": 36,
3167 |    "metadata": {},
3168 |    "outputs": [
3169 |     {
3170 |      "data": {
3171 |       "text/plain": [
3172 |        "28"
3173 |       ]
3174 |      },
3175 |      "execution_count": 36,
3176 |      "metadata": {},
3177 |      "output_type": "execute_result"
3178 |     }
3179 |    ],
3180 |    "source": [
3181 |     "# Get the number of unique companies in our data\n",
3182 |     "df.Company.nunique()"
3183 |    ]
3184 |   },
3185 |   {
3186 |    "cell_type": "code",
3187 |    "execution_count": 37,
3188 |    "metadata": {},
3189 |    "outputs": [
3190 |     {
3191 |      "data": {
3192 |       "text/plain": [
3193 |        "Amazon        3469\n",
3194 |        "Google        3445\n",
3195 |        "Facebook      1817\n",
3196 |        "Microsoft     1790\n",
3197 |        "IBM            873\n",
3198 |        "Cisco          787\n",
3199 |        "Oracle         701\n",
3200 |        "Uber           445\n",
3201 |        "Yelp           404\n",
3202 |        "Orange         363\n",
3203 |        "Intel          338\n",
3204 |        "Salesforce     313\n",
3205 |        "SAP            275\n",
3206 |        "Twitter        258\n",
3207 |        "Dell           258\n",
3208 |        "Airbnb         233\n",
3209 |        "NVIDIA         229\n",
3210 |        "Adobe          211\n",
3211 |        "Intuit         203\n",
3212 |        "PayPal         193\n",
3213 |        "Siemens        182\n",
3214 |        "Square         177\n",
3215 |        "Samsung        159\n",
3216 |        "eBay           148\n",
3217 |        "Symantec       147\n",
3218 |        "Snap           113\n",
3219 |        "Netflix        109\n",
3220 |        "Tencent         14\n",
3221 |        "Name: Company, dtype: int64"
3222 |       ]
3223 |      },
3224 |      "execution_count": 37,
3225 |      "metadata": {},
3226 |      "output_type": "execute_result"
3227 |     }
3228 |    ],
3229 |    "source": [
3230 |     "# You can also see how many reviews are for each company, sorted in a descending order.\n",
3231 |     "df.Company.value_counts()"
3232 |    ]
3233 |   },
3234 |   {
3235 |    "cell_type": "markdown",
3236 |    "metadata": {},
3237 |    "source": [
3238 |     "### 5.3 Plotting\n",
3239 |     "If you want to see the break down of process lengths for different companies, you can use `.plot()` with `.groupby()`.\n",
3240 |     "\n",
3241 |     "**Note**: Plotting in pandas is both mind-boggling and mind-blowing. If you're not familiar, you might want to check out some tutorials, e.g. [this simple tutorial](https://realpython.com/pandas-plot-python/) or [this saiyan-level pandas plotting with seaborn](https://jakevdp.github.io/PythonDataScienceHandbook/04.14-visualization-with-seaborn.html)."
3242 |    ]
3243 |   },
3244 |   {
3245 |    "cell_type": "code",
3246 |    "execution_count": 38,
3247 |    "metadata": {},
3248 |    "outputs": [
3249 |     {
3250 |      "data": {
3251 |       "text/plain": [
3252 |        "<matplotlib.axes._subplots.AxesSubplot at 0x7faac0d2ac88>"
3253 |       ]
3254 |      },
3255 |      "execution_count": 38,
3256 |      "metadata": {},
3257 |      "output_type": "execute_result"
3258 |     },
3259 |     {
3260 |      "data": {
3261 |       "image/png": "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\n",
3262 |       "text/plain": [
3263 |        "<Figure size 1080x576 with 1 Axes>"
3264 |       ]
3265 |      },
3266 |      "metadata": {
3267 |       "needs_background": "light"
3268 |      },
3269 |      "output_type": "display_data"
3270 |     }
3271 |    ],
3272 |    "source": [
3273 |     "# Group the DataFrame by \"Company\" and \"Process\", count the number of elements,\n",
3274 |     "# then unstack by \"Process\", then plot a bar chart\n",
3275 |     "df.groupby([\"Company\", \"Process\"]).size().unstack(level=1).plot(kind=\"bar\", figsize=(15, 8))"
3276 |    ]
3277 |   },
3278 |   {
3279 |    "cell_type": "markdown",
3280 |    "metadata": {},
3281 |    "source": [
3282 |     "## 6. Common pitfalls\n",
3283 |     "pandas is great for most day-to-day data analysis. It's instrumental to my job and I'm grateful that the entire pandas community is actively developing it. However, I think some of pandas design decisions are a bit questionable.\n",
3284 |     "\n",
3285 |     "Some of the common pandas pitfalls:\n",
3286 |     "### 6.1 NaNs\n",
3287 |     "NaNs are stored as floats in pandas, so when an operation fails because of NaNs, it doesn't say that there's a NaN but because that operation doesn't exist for floats.\n",
3288 |     "\n",
3289 |     "### 6.2 Changes not Inplace \n",
3290 |     "Most pandas operations aren't inplace by default, so if you make changes to your `DataFrame`, you need to assign the changes back to your DataFrame. You can make changes inplace by setting argument `inplace=True`."
3291 |    ]
3292 |   },
3293 |   {
3294 |    "cell_type": "code",
3295 |    "execution_count": 39,
3296 |    "metadata": {},
3297 |    "outputs": [
3298 |     {
3299 |      "data": {
3300 |       "text/plain": [
3301 |        "Index(['Company', 'Title', 'Job', 'Level', 'Date', 'Upvotes', 'Offer',\n",
3302 |        "       'Experience', 'Difficulty', 'Review', 'Process'],\n",
3303 |        "      dtype='object')"
3304 |       ]
3305 |      },
3306 |      "execution_count": 39,
3307 |      "metadata": {},
3308 |      "output_type": "execute_result"
3309 |     }
3310 |    ],
3311 |    "source": [
3312 |     "# \"Process\" column is still in df\n",
3313 |     "df.drop(columns=[\"Process\"])\n",
3314 |     "df.columns"
3315 |    ]
3316 |   },
3317 |   {
3318 |    "cell_type": "code",
3319 |    "execution_count": 40,
3320 |    "metadata": {},
3321 |    "outputs": [
3322 |     {
3323 |      "data": {
3324 |       "text/plain": [
3325 |        "Index(['Company', 'Title', 'Job', 'Level', 'Date', 'Upvotes', 'Offer',\n",
3326 |        "       'Experience', 'Difficulty', 'Review'],\n",
3327 |        "      dtype='object')"
3328 |       ]
3329 |      },
3330 |      "execution_count": 40,
3331 |      "metadata": {},
3332 |      "output_type": "execute_result"
3333 |     }
3334 |    ],
3335 |    "source": [
3336 |     "# To make changes to df, set `inplace=True`\n",
3337 |     "df.drop(columns=[\"Process\"], inplace=True)\n",
3338 |     "df.columns\n",
3339 |     "# This is equivalent to\n",
3340 |     "# df = df.drop(columns=[\"Process\"])"
3341 |    ]
3342 |   },
3343 |   {
3344 |    "cell_type": "markdown",
3345 |    "metadata": {},
3346 |    "source": [
3347 |     "### 6.3 Performance issues with very large datasets\n",
3348 |     "\n",
3349 |     "### 6.4 Reproducibility issues\n",
3350 |     "Especially with dumping and loading `DataFrame` to/from files. There are two main causes:\n",
3351 |     "\n",
3352 |     "- Problem with labels (see the section about labels above).\n",
3353 |     "- [Weird rounding issues for floats](https://stackoverflow.com/questions/47368296/pandas-read-csv-file-with-float-values-results-in-weird-rounding-and-decimal-dig). \n",
3354 |     "\n",
3355 |     "### 6.5 Not GPU compatible\n",
3356 |     "pandas can't take advantage of GPUs, so if your computations are on on GPUs and your feature engineering is on CPUs, it can become a time bottleneck to move data from CPUs to GPUs. If you want something like pandas but works on GPUs, check out dask and modin."
3357 |    ]
3358 |   },
3359 |   {
3360 |    "cell_type": "markdown",
3361 |    "metadata": {},
3362 |    "source": [
3363 |     "Oh, and this is [the analysis I did based on this data](https://huyenchip.com/2019/08/21/glassdoor-interview-reviews-tech-hiring-cultures.html), in case you're interested!"
3364 |    ]
3365 |   }
3366 |  ],
3367 |  "metadata": {
3368 |   "kernelspec": {
3369 |    "display_name": "Python 3",
3370 |    "language": "python",
3371 |    "name": "python3"
3372 |   },
3373 |   "language_info": {
3374 |    "codemirror_mode": {
3375 |     "name": "ipython",
3376 |     "version": 3
3377 |    },
3378 |    "file_extension": ".py",
3379 |    "mimetype": "text/x-python",
3380 |    "name": "python",
3381 |    "nbconvert_exporter": "python",
3382 |    "pygments_lexer": "ipython3",
3383 |    "version": "3.6.10"
3384 |   },
3385 |   "varInspector": {
3386 |    "cols": {
3387 |     "lenName": 16,
3388 |     "lenType": 16,
3389 |     "lenVar": 40
3390 |    },
3391 |    "kernels_config": {
3392 |     "python": {
3393 |      "delete_cmd_postfix": "",
3394 |      "delete_cmd_prefix": "del ",
3395 |      "library": "var_list.py",
3396 |      "varRefreshCmd": "print(var_dic_list())"
3397 |     },
3398 |     "r": {
3399 |      "delete_cmd_postfix": ") ",
3400 |      "delete_cmd_prefix": "rm(",
3401 |      "library": "var_list.r",
3402 |      "varRefreshCmd": "cat(var_dic_list()) "
3403 |     }
3404 |    },
3405 |    "types_to_exclude": [
3406 |     "module",
3407 |     "function",
3408 |     "builtin_function_or_method",
3409 |     "instance",
3410 |     "_Feature"
3411 |    ],
3412 |    "window_display": false
3413 |   }
3414 |  },
3415 |  "nbformat": 4,
3416 |  "nbformat_minor": 2
3417 | }
3418 | 


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