├── AmesHousing.tsv ├── House_Sale_Price_Prediction.ipynb └── README.md /House_Sale_Price_Prediction.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": { 6 | "collapsed": true 7 | }, 8 | "source": [ 9 | "# Predicting House Sale Prices" 10 | ] 11 | }, 12 | { 13 | "cell_type": "markdown", 14 | "metadata": {}, 15 | "source": [ 16 | "## Import Necessary Libraries" 17 | ] 18 | }, 19 | { 20 | "cell_type": "code", 21 | "execution_count": 350, 22 | "metadata": {}, 23 | "outputs": [], 24 | "source": [ 25 | "import pandas as pd\n", 26 | "import numpy as np\n", 27 | "import matplotlib.pyplot as plt\n", 28 | "from sklearn.model_selection import KFold\n", 29 | "from sklearn.metrics import mean_squared_error\n", 30 | "from sklearn import linear_model" 31 | ] 32 | }, 33 | { 34 | "cell_type": "markdown", 35 | "metadata": {}, 36 | "source": [ 37 | "## Housing Dataset" 38 | ] 39 | }, 40 | { 41 | "cell_type": "markdown", 42 | "metadata": {}, 43 | "source": [ 44 | "Here is the housing data for the city of Ames, Iowa, United States from 2006 to 2010. " 45 | ] 46 | }, 47 | { 48 | "cell_type": "code", 49 | "execution_count": 351, 50 | "metadata": {}, 51 | "outputs": [ 52 | { 53 | "data": { 54 | "text/html": [ 55 | "
\n", 56 | "\n", 69 | "\n", 70 | " \n", 71 | " \n", 72 | " \n", 73 | " \n", 74 | " \n", 75 | " \n", 76 | " \n", 77 | " \n", 78 | " \n", 79 | " \n", 80 | " \n", 81 | " \n", 82 | " \n", 83 | " \n", 84 | " \n", 85 | " \n", 86 | " \n", 87 | " \n", 88 | " \n", 89 | " \n", 90 | " \n", 91 | " \n", 92 | " \n", 93 | " \n", 94 | " \n", 95 | " \n", 96 | " \n", 97 | " \n", 98 | " \n", 99 | " \n", 100 | " \n", 101 | " \n", 102 | " \n", 103 | " \n", 104 | " \n", 105 | " \n", 106 | " \n", 107 | " \n", 108 | " \n", 109 | " \n", 110 | " \n", 111 | " \n", 112 | " \n", 113 | " \n", 114 | " \n", 115 | " \n", 116 | " \n", 117 | " \n", 118 | " \n", 119 | " \n", 120 | " \n", 121 | " \n", 122 | " \n", 123 | " \n", 124 | " \n", 125 | " \n", 126 | " \n", 127 | " \n", 128 | " \n", 129 | " \n", 130 | " \n", 131 | " \n", 132 | " \n", 133 | " \n", 134 | " \n", 135 | " \n", 136 | " \n", 137 | " \n", 138 | " \n", 139 | " \n", 140 | " \n", 141 | " \n", 142 | " \n", 143 | " \n", 144 | " \n", 145 | " \n", 146 | " \n", 147 | " \n", 148 | " \n", 149 | " \n", 150 | " \n", 151 | " \n", 152 | " \n", 153 | " \n", 154 | " \n", 155 | " \n", 156 | " \n", 157 | " \n", 158 | " \n", 159 | " \n", 160 | " \n", 161 | " \n", 162 | " \n", 163 | " \n", 164 | " \n", 165 | " \n", 166 | " \n", 167 | " \n", 168 | " \n", 169 | " \n", 170 | " \n", 171 | " \n", 172 | " \n", 173 | " \n", 174 | " \n", 175 | " \n", 176 | " \n", 177 | " \n", 178 | " \n", 179 | " \n", 180 | " \n", 181 | " \n", 182 | " \n", 183 | " \n", 184 | " \n", 185 | " \n", 186 | " \n", 187 | " \n", 188 | " \n", 189 | " \n", 190 | " \n", 191 | " \n", 192 | " \n", 193 | " \n", 194 | " \n", 195 | " \n", 196 | " \n", 197 | " \n", 198 | " \n", 199 | " \n", 200 | " \n", 201 | " \n", 202 | " \n", 203 | " \n", 204 | " \n", 205 | " \n", 206 | " \n", 207 | " \n", 208 | " \n", 209 | " \n", 210 | " \n", 211 | " \n", 212 | " \n", 213 | " \n", 214 | " \n", 215 | " \n", 216 | " \n", 217 | " \n", 218 | " \n", 219 | " \n", 220 | " \n", 221 | " \n", 222 | " \n", 223 | " \n", 224 | " \n", 225 | " \n", 226 | " \n", 227 | " \n", 228 | " \n", 229 | " \n", 230 | " \n", 231 | " \n", 232 | " \n", 233 | " \n", 234 | " \n", 235 | " \n", 236 | " \n", 237 | " \n", 238 | " \n", 239 | " \n", 240 | " \n", 241 | " \n", 242 | " \n", 243 | " \n", 244 | " \n", 245 | " \n", 246 | " \n", 247 | " \n", 248 | " \n", 249 | " \n", 250 | " \n", 251 | " \n", 252 | " \n", 253 | " \n", 254 | " \n", 255 | " \n", 256 | " \n", 257 | " \n", 258 | " \n", 259 | " \n", 260 | " \n", 261 | " \n", 262 | " \n", 263 | " \n", 264 | " \n", 265 | " \n", 266 | " \n", 267 | " \n", 268 | " \n", 269 | " \n", 270 | " \n", 271 | " \n", 272 | " \n", 273 | " \n", 274 | " \n", 275 | " \n", 276 | " \n", 277 | " \n", 278 | " \n", 279 | " \n", 280 | " \n", 281 | " \n", 282 | " \n", 283 | " \n", 284 | " \n", 285 | " \n", 286 | " \n", 287 | " \n", 288 | " \n", 289 | " \n", 290 | " \n", 291 | " \n", 292 | " \n", 293 | " \n", 294 | " \n", 295 | " \n", 296 | " \n", 297 | " \n", 298 | " \n", 299 | " \n", 300 | " \n", 301 | " \n", 302 | " \n", 303 | " \n", 304 | " \n", 305 | " \n", 306 | " \n", 307 | " \n", 308 | " \n", 309 | " \n", 310 | " \n", 311 | " \n", 312 | " \n", 313 | " \n", 314 | " \n", 315 | " \n", 316 | " \n", 317 | " \n", 318 | " \n", 319 | " \n", 320 | " \n", 321 | " \n", 322 | " \n", 323 | " \n", 324 | " \n", 325 | " \n", 326 | " \n", 327 | " \n", 328 | " \n", 329 | " \n", 330 | " \n", 331 | " \n", 332 | " \n", 333 | " \n", 334 | " \n", 335 | " \n", 336 | " \n", 337 | " \n", 338 | " \n", 339 | " \n", 340 | " \n", 341 | " \n", 342 | " \n", 343 | " \n", 344 | " \n", 345 | " \n", 346 | " \n", 347 | " \n", 348 | " \n", 349 | " \n", 350 | " \n", 351 | " \n", 352 | " \n", 353 | " \n", 354 | " \n", 355 | " \n", 356 | " \n", 357 | " \n", 358 | " \n", 359 | " \n", 360 | " \n", 361 | " \n", 362 | "
OrderPIDMS SubClassMS ZoningLot FrontageLot AreaStreetAlleyLot ShapeLand Contour...Pool AreaPool QCFenceMisc FeatureMisc ValMo SoldYr SoldSale TypeSale ConditionSalePrice
0152630110020RL141.031770PaveNaNIR1Lvl...0NaNNaNNaN052010WDNormal215000
1252635004020RH80.011622PaveNaNRegLvl...0NaNMnPrvNaN062010WDNormal105000
2352635101020RL81.014267PaveNaNIR1Lvl...0NaNNaNGar21250062010WDNormal172000
3452635303020RL93.011160PaveNaNRegLvl...0NaNNaNNaN042010WDNormal244000
4552710501060RL74.013830PaveNaNIR1Lvl...0NaNMnPrvNaN032010WDNormal189900
..................................................................
2925292692327508080RL37.07937PaveNaNIR1Lvl...0NaNGdPrvNaN032006WDNormal142500
2926292792327610020RLNaN8885PaveNaNIR1Low...0NaNMnPrvNaN062006WDNormal131000
2927292892340012585RL62.010441PaveNaNRegLvl...0NaNMnPrvShed70072006WDNormal132000
2928292992410007020RL77.010010PaveNaNRegLvl...0NaNNaNNaN042006WDNormal170000
2929293092415105060RL74.09627PaveNaNRegLvl...0NaNNaNNaN0112006WDNormal188000
\n", 363 | "

2930 rows × 82 columns

\n", 364 | "
" 365 | ], 366 | "text/plain": [ 367 | " Order PID MS SubClass MS Zoning Lot Frontage Lot Area Street \\\n", 368 | "0 1 526301100 20 RL 141.0 31770 Pave \n", 369 | "1 2 526350040 20 RH 80.0 11622 Pave \n", 370 | "2 3 526351010 20 RL 81.0 14267 Pave \n", 371 | "3 4 526353030 20 RL 93.0 11160 Pave \n", 372 | "4 5 527105010 60 RL 74.0 13830 Pave \n", 373 | "... ... ... ... ... ... ... ... \n", 374 | "2925 2926 923275080 80 RL 37.0 7937 Pave \n", 375 | "2926 2927 923276100 20 RL NaN 8885 Pave \n", 376 | "2927 2928 923400125 85 RL 62.0 10441 Pave \n", 377 | "2928 2929 924100070 20 RL 77.0 10010 Pave \n", 378 | "2929 2930 924151050 60 RL 74.0 9627 Pave \n", 379 | "\n", 380 | " Alley Lot Shape Land Contour ... Pool Area Pool QC Fence Misc Feature \\\n", 381 | "0 NaN IR1 Lvl ... 0 NaN NaN NaN \n", 382 | "1 NaN Reg Lvl ... 0 NaN MnPrv NaN \n", 383 | "2 NaN IR1 Lvl ... 0 NaN NaN Gar2 \n", 384 | "3 NaN Reg Lvl ... 0 NaN NaN NaN \n", 385 | "4 NaN IR1 Lvl ... 0 NaN MnPrv NaN \n", 386 | "... ... ... ... ... ... ... ... ... \n", 387 | "2925 NaN IR1 Lvl ... 0 NaN GdPrv NaN \n", 388 | "2926 NaN IR1 Low ... 0 NaN MnPrv NaN \n", 389 | "2927 NaN Reg Lvl ... 0 NaN MnPrv Shed \n", 390 | "2928 NaN Reg Lvl ... 0 NaN NaN NaN \n", 391 | "2929 NaN Reg Lvl ... 0 NaN NaN NaN \n", 392 | "\n", 393 | " Misc Val Mo Sold Yr Sold Sale Type Sale Condition SalePrice \n", 394 | "0 0 5 2010 WD Normal 215000 \n", 395 | "1 0 6 2010 WD Normal 105000 \n", 396 | "2 12500 6 2010 WD Normal 172000 \n", 397 | "3 0 4 2010 WD Normal 244000 \n", 398 | "4 0 3 2010 WD Normal 189900 \n", 399 | "... ... ... ... ... ... ... \n", 400 | "2925 0 3 2006 WD Normal 142500 \n", 401 | "2926 0 6 2006 WD Normal 131000 \n", 402 | "2927 700 7 2006 WD Normal 132000 \n", 403 | "2928 0 4 2006 WD Normal 170000 \n", 404 | "2929 0 11 2006 WD Normal 188000 \n", 405 | "\n", 406 | "[2930 rows x 82 columns]" 407 | ] 408 | }, 409 | "execution_count": 351, 410 | "metadata": {}, 411 | "output_type": "execute_result" 412 | } 413 | ], 414 | "source": [ 415 | "df = pd.read_csv(\"AmesHousing.tsv\", sep=\"\\t\")\n", 416 | "df" 417 | ] 418 | }, 419 | { 420 | "cell_type": "markdown", 421 | "metadata": {}, 422 | "source": [ 423 | "## Feature Engineering, Data Exploration, and Data Cleaning" 424 | ] 425 | }, 426 | { 427 | "cell_type": "markdown", 428 | "metadata": {}, 429 | "source": [ 430 | "Steps:\n", 431 | "1. See which of the features have missing values. For the features that has more than 5% of values missing, drop the features.\n", 432 | "2. For the text features, drop those that have missing values.\n", 433 | "3. For the numeric features that have up to 5% missing values, replace the missing values with the most occurring value (i.e mode).\n", 434 | "4. Remove the columns that leak about the final sale and columns useless for building models." 435 | ] 436 | }, 437 | { 438 | "cell_type": "code", 439 | "execution_count": 359, 440 | "metadata": {}, 441 | "outputs": [], 442 | "source": [ 443 | "def transform_features(df):\n", 444 | " \n", 445 | " feature_data = {\"Feature\": df.isnull().sum().keys().tolist(), \"Missing Values\": df.isnull().sum().values.tolist()}\n", 446 | " df_features = pd.DataFrame(feature_data)\n", 447 | " rows,cols = df.shape\n", 448 | " df_features['Percent Missing'] = 100*df_features['Missing Values']/rows\n", 449 | " features = df_features[df_features['Percent Missing'] > 5]['Feature'].tolist()\n", 450 | " df = df.drop(features, axis=1)\n", 451 | " \n", 452 | " text_feature_data = {\"Feature\": df.select_dtypes(include=['object']).isnull().sum().keys().tolist(), \"Missing Values\": df.select_dtypes(include=['object']).isnull().sum().values.tolist()}\n", 453 | " df_text_features = pd.DataFrame(text_feature_data)\n", 454 | " text_features = df_text_features[df_text_features['Missing Values'] > 0]['Feature'].tolist()\n", 455 | " df = df.drop(text_features, axis=1)\n", 456 | " \n", 457 | " numeric_dtypes_missing_values = df.select_dtypes(include=['int64','float64']).isnull().sum()\n", 458 | " numeric_features = numeric_dtypes_missing_values[numeric_dtypes_missing_values > 0]\n", 459 | " df = df.fillna(value=numeric_features)\n", 460 | " \n", 461 | " df['Years Since Remod'] = df['Yr Sold'] - df['Year Remod/Add']\n", 462 | " df['Years Before Sale'] = df['Yr Sold'] - df['Year Built']\n", 463 | " df = df.drop(['Year Built', 'Year Remod/Add'], axis=1)\n", 464 | " df = df.drop([1702,2180,2181], axis=0)\n", 465 | " df = df.drop([\"Order\",\"PID\",\"Mo Sold\", \"Sale Condition\", \"Sale Type\", \"Yr Sold\"], axis=1)\n", 466 | " \n", 467 | " return df" 468 | ] 469 | }, 470 | { 471 | "cell_type": "markdown", 472 | "metadata": {}, 473 | "source": [ 474 | "## Feature Selection" 475 | ] 476 | }, 477 | { 478 | "cell_type": "markdown", 479 | "metadata": {}, 480 | "source": [ 481 | "Steps:\n", 482 | "1. Generate a correlation heatmap matrix of the numerical features in the training data set with respect to SalePrice.\n", 483 | "2. Determine which numerical features should be converted to categorical features as their numbers don't have any semantic meaning. Aim for the features with few unique values with a significant majority (95%) of the values belonging to a specific category." 484 | ] 485 | }, 486 | { 487 | "cell_type": "code", 488 | "execution_count": 360, 489 | "metadata": {}, 490 | "outputs": [], 491 | "source": [ 492 | "def select_features(df, coeff_threshold=0.4, uniq_threshold=10):\n", 493 | " numerical_df = transform_df.select_dtypes(include=['int64', 'float64'])\n", 494 | " abs_corr_coeffs = numerical_df.corr()['SalePrice'].abs().sort_values()\n", 495 | " df = df.drop(abs_corr_coeffs[abs_corr_coeffs < coeff_threshold].index, axis=1)\n", 496 | " \n", 497 | " nominal_features = [\"PID\", \"MS SubClass\", \"MS Zoning\", \"Street\", \"Alley\", \"Land Contour\", \"Lot Config\", \"Neighborhood\", \n", 498 | " \"Condition 1\", \"Condition 2\", \"Bldg Type\", \"House Style\", \"Roof Style\", \"Roof Matl\", \"Exterior 1st\", \n", 499 | " \"Exterior 2nd\", \"Mas Vnr Type\", \"Foundation\", \"Heating\", \"Central Air\", \"Garage Type\", \n", 500 | " \"Misc Feature\", \"Sale Type\", \"Sale Condition\"]\n", 501 | " \n", 502 | " transform_cat_cols = []\n", 503 | " for col in nominal_features:\n", 504 | " if col in df.columns:\n", 505 | " transform_cat_cols.append(col)\n", 506 | "\n", 507 | " uniqueness_counts = df[transform_cat_cols].apply(lambda col: len(col.value_counts())).sort_values()\n", 508 | " drop_nonuniq_cols = uniqueness_counts[uniqueness_counts > 10].index\n", 509 | " df = df.drop(drop_nonuniq_cols, axis=1)\n", 510 | " \n", 511 | " text_cols = df.select_dtypes(include=['object'])\n", 512 | " for col in text_cols:\n", 513 | " df[col] = df[col].astype('category')\n", 514 | " df = pd.concat([df, pd.get_dummies(df.select_dtypes(include=['category']))], axis=1).drop(text_cols,axis=1)\n", 515 | " \n", 516 | " return df" 517 | ] 518 | }, 519 | { 520 | "cell_type": "markdown", 521 | "metadata": {}, 522 | "source": [ 523 | "## Train and Test the Model" 524 | ] 525 | }, 526 | { 527 | "cell_type": "markdown", 528 | "metadata": {}, 529 | "source": [ 530 | "The function train_test can perform validation (k=0) or K-Fold cross validation (k > 0). " 531 | ] 532 | }, 533 | { 534 | "cell_type": "code", 535 | "execution_count": 361, 536 | "metadata": {}, 537 | "outputs": [], 538 | "source": [ 539 | "def train_test(df, k=0):\n", 540 | " if k < 0:\n", 541 | " raise Exception(\"Value for k is negative!\")\n", 542 | " \n", 543 | " numeric_df = df.select_dtypes(include=['integer', 'float'])\n", 544 | " features = numeric_df.columns.drop(\"SalePrice\")\n", 545 | " lr = linear_model.LinearRegression()\n", 546 | " \n", 547 | " if k==0:\n", 548 | " row,col = df.shape\n", 549 | " train_test_cutoff = row*0.7\n", 550 | " train = numeric_df.iloc[:int(train_test_cutoff)]\n", 551 | " test = numeric_df.iloc[int(train_test_cutoff):]\n", 552 | " lr.fit(train[features], train[\"SalePrice\"])\n", 553 | " predictions = lr.predict(test[features])\n", 554 | " mse = mean_squared_error(test[\"SalePrice\"], predictions)\n", 555 | " \n", 556 | " return np.sqrt(mse)\n", 557 | " \n", 558 | " else:\n", 559 | " kf = KFold(n_splits=k, shuffle=True)\n", 560 | " rmse_values = []\n", 561 | " for train_index, test_index, in kf.split(df):\n", 562 | " train = df.iloc[train_index]\n", 563 | " test = df.iloc[test_index]\n", 564 | " lr.fit(train[features], train[\"SalePrice\"])\n", 565 | " predictions = lr.predict(test[features])\n", 566 | " mse = mean_squared_error(test[\"SalePrice\"], predictions)\n", 567 | " rmse = np.sqrt(mse)\n", 568 | " rmse_values.append(rmse)\n", 569 | " print(rmse_values)\n", 570 | " avg_rmse = np.mean(rmse_values)\n", 571 | " return avg_rmse\n", 572 | " " 573 | ] 574 | }, 575 | { 576 | "cell_type": "markdown", 577 | "metadata": {}, 578 | "source": [ 579 | "## Complete Automated ML Pipeline" 580 | ] 581 | }, 582 | { 583 | "cell_type": "code", 584 | "execution_count": 362, 585 | "metadata": {}, 586 | "outputs": [ 587 | { 588 | "name": "stdout", 589 | "output_type": "stream", 590 | "text": [ 591 | "[24344.031226139265, 29103.803299408657, 35693.48786672571, 26965.662593750727]\n" 592 | ] 593 | }, 594 | { 595 | "data": { 596 | "text/plain": [ 597 | "29026.74624650609" 598 | ] 599 | }, 600 | "execution_count": 362, 601 | "metadata": {}, 602 | "output_type": "execute_result" 603 | } 604 | ], 605 | "source": [ 606 | "transform_df = transform_features(df)\n", 607 | "filtered_df = select_features(transform_df)\n", 608 | "rmse = train_test(filtered_df, k=4)\n", 609 | "rmse" 610 | ] 611 | }, 612 | { 613 | "cell_type": "code", 614 | "execution_count": null, 615 | "metadata": {}, 616 | "outputs": [], 617 | "source": [] 618 | } 619 | ], 620 | "metadata": { 621 | "anaconda-cloud": {}, 622 | "kernelspec": { 623 | "display_name": "Python 3", 624 | "language": "python", 625 | "name": "python3" 626 | }, 627 | "language_info": { 628 | "codemirror_mode": { 629 | "name": "ipython", 630 | "version": 3 631 | }, 632 | "file_extension": ".py", 633 | "mimetype": "text/x-python", 634 | "name": "python", 635 | "nbconvert_exporter": "python", 636 | "pygments_lexer": "ipython3", 637 | "version": "3.6.8" 638 | } 639 | }, 640 | "nbformat": 4, 641 | "nbformat_minor": 2 642 | } 643 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Project: House Sale Price Prediction 2 | 3 | ## Project Goal 4 | Investigate housing data set and build an ML Pipeline that automates feature engineering, feature selection, and training and testing models for estimation and prediction of housing prices. 5 | 6 | ## Dataset Information: 7 | AmesHousing.tsv is a data set describing the sale of individual residential property in Ames, Iowa from 2006 to 2010. The data set contains 2930 observations and a large number of explanatory variables (23 nominal, 23 ordinal, 14 discrete, and 20 continuous) involved in assessing home values. 8 | 9 | ## Tech Stack 10 | Python
11 | Jupyter Notebook
12 | NumPy
13 | Pandas
14 | Matplotlib
15 | Scikit-learn
16 | 17 | ## Featured ML Algorithms 18 | K-Fold Cross Validation
19 | Linear Regression
--------------------------------------------------------------------------------