├── .gitignore
├── 01-Constant-features.ipynb
├── 02-One-way-ANOVA.ipynb
├── 03-Lasso.ipynb
├── 04-Feature-shuffling.ipynb
├── 2022_DataTalksClub_FeatureSelection.pdf
├── LICENSE
└── README.md


/.gitignore:
--------------------------------------------------------------------------------
1 | # Jupyter Notebook
2 | .ipynb_checkpoints
3 | 
4 | # datasets
5 | *.csv
6 | *.zip


--------------------------------------------------------------------------------
/01-Constant-features.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "metadata": {},
  6 |    "source": [
  7 |     "## Constant features"
  8 |    ]
  9 |   },
 10 |   {
 11 |    "cell_type": "code",
 12 |    "execution_count": 1,
 13 |    "metadata": {},
 14 |    "outputs": [],
 15 |    "source": [
 16 |     "import pandas as pd\n",
 17 |     "\n",
 18 |     "from sklearn.datasets import make_classification\n",
 19 |     "from sklearn.model_selection import train_test_split\n",
 20 |     "\n",
 21 |     "from sklearn.feature_selection import VarianceThreshold\n",
 22 |     "from feature_engine.selection import DropConstantFeatures"
 23 |    ]
 24 |   },
 25 |   {
 26 |    "cell_type": "code",
 27 |    "execution_count": 2,
 28 |    "metadata": {},
 29 |    "outputs": [
 30 |     {
 31 |      "data": {
 32 |       "text/html": [
 33 |        "<div>\n",
 34 |        "<style scoped>\n",
 35 |        "    .dataframe tbody tr th:only-of-type {\n",
 36 |        "        vertical-align: middle;\n",
 37 |        "    }\n",
 38 |        "\n",
 39 |        "    .dataframe tbody tr th {\n",
 40 |        "        vertical-align: top;\n",
 41 |        "    }\n",
 42 |        "\n",
 43 |        "    .dataframe thead th {\n",
 44 |        "        text-align: right;\n",
 45 |        "    }\n",
 46 |        "</style>\n",
 47 |        "<table border=\"1\" class=\"dataframe\">\n",
 48 |        "  <thead>\n",
 49 |        "    <tr style=\"text-align: right;\">\n",
 50 |        "      <th></th>\n",
 51 |        "      <th>0</th>\n",
 52 |        "      <th>1</th>\n",
 53 |        "      <th>2</th>\n",
 54 |        "      <th>3</th>\n",
 55 |        "      <th>4</th>\n",
 56 |        "      <th>5</th>\n",
 57 |        "      <th>6</th>\n",
 58 |        "      <th>7</th>\n",
 59 |        "      <th>8</th>\n",
 60 |        "      <th>9</th>\n",
 61 |        "    </tr>\n",
 62 |        "  </thead>\n",
 63 |        "  <tbody>\n",
 64 |        "    <tr>\n",
 65 |        "      <th>0</th>\n",
 66 |        "      <td>1</td>\n",
 67 |        "      <td>-0.376539</td>\n",
 68 |        "      <td>-0.620180</td>\n",
 69 |        "      <td>-0.157567</td>\n",
 70 |        "      <td>-1.120805</td>\n",
 71 |        "      <td>1</td>\n",
 72 |        "      <td>-1.574578</td>\n",
 73 |        "      <td>1.678046</td>\n",
 74 |        "      <td>1.080180</td>\n",
 75 |        "      <td>1</td>\n",
 76 |        "    </tr>\n",
 77 |        "    <tr>\n",
 78 |        "      <th>1</th>\n",
 79 |        "      <td>1</td>\n",
 80 |        "      <td>0.762409</td>\n",
 81 |        "      <td>-0.784210</td>\n",
 82 |        "      <td>-0.096479</td>\n",
 83 |        "      <td>-0.408758</td>\n",
 84 |        "      <td>1</td>\n",
 85 |        "      <td>0.210942</td>\n",
 86 |        "      <td>-0.850449</td>\n",
 87 |        "      <td>-0.461301</td>\n",
 88 |        "      <td>1</td>\n",
 89 |        "    </tr>\n",
 90 |        "    <tr>\n",
 91 |        "      <th>2</th>\n",
 92 |        "      <td>1</td>\n",
 93 |        "      <td>2.227934</td>\n",
 94 |        "      <td>0.547727</td>\n",
 95 |        "      <td>-0.341481</td>\n",
 96 |        "      <td>-0.817577</td>\n",
 97 |        "      <td>1</td>\n",
 98 |        "      <td>-2.663678</td>\n",
 99 |        "      <td>2.440042</td>\n",
100 |        "      <td>1.698919</td>\n",
101 |        "      <td>1</td>\n",
102 |        "    </tr>\n",
103 |        "    <tr>\n",
104 |        "      <th>3</th>\n",
105 |        "      <td>1</td>\n",
106 |        "      <td>0.061129</td>\n",
107 |        "      <td>-0.995868</td>\n",
108 |        "      <td>-0.214351</td>\n",
109 |        "      <td>-0.558957</td>\n",
110 |        "      <td>1</td>\n",
111 |        "      <td>-2.149167</td>\n",
112 |        "      <td>2.294192</td>\n",
113 |        "      <td>-1.383965</td>\n",
114 |        "      <td>1</td>\n",
115 |        "    </tr>\n",
116 |        "    <tr>\n",
117 |        "      <th>4</th>\n",
118 |        "      <td>1</td>\n",
119 |        "      <td>0.046349</td>\n",
120 |        "      <td>0.834756</td>\n",
121 |        "      <td>-0.104845</td>\n",
122 |        "      <td>-0.455528</td>\n",
123 |        "      <td>1</td>\n",
124 |        "      <td>-0.911018</td>\n",
125 |        "      <td>0.898098</td>\n",
126 |        "      <td>1.068259</td>\n",
127 |        "      <td>1</td>\n",
128 |        "    </tr>\n",
129 |        "  </tbody>\n",
130 |        "</table>\n",
131 |        "</div>"
132 |       ],
133 |       "text/plain": [
134 |        "   0         1         2         3         4  5         6         7         8  \\\n",
135 |        "0  1 -0.376539 -0.620180 -0.157567 -1.120805  1 -1.574578  1.678046  1.080180   \n",
136 |        "1  1  0.762409 -0.784210 -0.096479 -0.408758  1  0.210942 -0.850449 -0.461301   \n",
137 |        "2  1  2.227934  0.547727 -0.341481 -0.817577  1 -2.663678  2.440042  1.698919   \n",
138 |        "3  1  0.061129 -0.995868 -0.214351 -0.558957  1 -2.149167  2.294192 -1.383965   \n",
139 |        "4  1  0.046349  0.834756 -0.104845 -0.455528  1 -0.911018  0.898098  1.068259   \n",
140 |        "\n",
141 |        "   9  \n",
142 |        "0  1  \n",
143 |        "1  1  \n",
144 |        "2  1  \n",
145 |        "3  1  \n",
146 |        "4  1  "
147 |       ]
148 |      },
149 |      "execution_count": 2,
150 |      "metadata": {},
151 |      "output_type": "execute_result"
152 |     }
153 |    ],
154 |    "source": [
155 |     "# Toy dataset with redundant and constant features\n",
156 |     "\n",
157 |     "X, y = make_classification(\n",
158 |     "    n_samples=1000,\n",
159 |     "    n_features=10,\n",
160 |     "    n_classes=2,\n",
161 |     "    random_state=10,\n",
162 |     ")\n",
163 |     "\n",
164 |     "X = pd.DataFrame(X)\n",
165 |     "y = pd.Series(y)\n",
166 |     "\n",
167 |     "# Add constant features\n",
168 |     "X[[0, 5, 9]] = 1\n",
169 |     "\n",
170 |     "X.head()"
171 |    ]
172 |   },
173 |   {
174 |    "cell_type": "code",
175 |    "execution_count": 3,
176 |    "metadata": {},
177 |    "outputs": [
178 |     {
179 |      "data": {
180 |       "text/plain": [
181 |        "((700, 10), (300, 10))"
182 |       ]
183 |      },
184 |      "execution_count": 3,
185 |      "metadata": {},
186 |      "output_type": "execute_result"
187 |     }
188 |    ],
189 |    "source": [
190 |     "# separate dataset into train and test\n",
191 |     "\n",
192 |     "X_train, X_test, y_train, y_test = train_test_split(\n",
193 |     "    X, y, test_size=0.3, random_state=0,\n",
194 |     ")\n",
195 |     "\n",
196 |     "X_train.shape, X_test.shape"
197 |    ]
198 |   },
199 |   {
200 |    "cell_type": "markdown",
201 |    "metadata": {},
202 |    "source": [
203 |     "## VarianceThreshold from Scikit-learn\n",
204 |     "\n",
205 |     "Only works with numerical variables. Categorical variables need to be encoded first."
206 |    ]
207 |   },
208 |   {
209 |    "cell_type": "code",
210 |    "execution_count": 4,
211 |    "metadata": {},
212 |    "outputs": [
213 |     {
214 |      "data": {
215 |       "text/html": [
216 |        "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>VarianceThreshold(threshold=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">VarianceThreshold</label><div class=\"sk-toggleable__content\"><pre>VarianceThreshold(threshold=0)</pre></div></div></div></div></div>"
217 |       ],
218 |       "text/plain": [
219 |        "VarianceThreshold(threshold=0)"
220 |       ]
221 |      },
222 |      "execution_count": 4,
223 |      "metadata": {},
224 |      "output_type": "execute_result"
225 |     }
226 |    ],
227 |    "source": [
228 |     "# To remove constant features\n",
229 |     "sel = VarianceThreshold(threshold=0)\n",
230 |     "\n",
231 |     "# fit finds the features with zero variance\n",
232 |     "sel.fit(X_train)  "
233 |    ]
234 |   },
235 |   {
236 |    "cell_type": "code",
237 |    "execution_count": 5,
238 |    "metadata": {},
239 |    "outputs": [
240 |     {
241 |      "data": {
242 |       "text/plain": [
243 |        "7"
244 |       ]
245 |      },
246 |      "execution_count": 5,
247 |      "metadata": {},
248 |      "output_type": "execute_result"
249 |     }
250 |    ],
251 |    "source": [
252 |     "# get_support is a boolean vector flags \n",
253 |     "# the features to keep\n",
254 |     "\n",
255 |     "# Number of selected features (the non-constant)\n",
256 |     "\n",
257 |     "sum(sel.get_support())"
258 |    ]
259 |   },
260 |   {
261 |    "cell_type": "code",
262 |    "execution_count": 6,
263 |    "metadata": {},
264 |    "outputs": [
265 |     {
266 |      "data": {
267 |       "text/plain": [
268 |        "Int64Index([0, 5, 9], dtype='int64')"
269 |       ]
270 |      },
271 |      "execution_count": 6,
272 |      "metadata": {},
273 |      "output_type": "execute_result"
274 |     }
275 |    ],
276 |    "source": [
277 |     "# the constant feautures\n",
278 |     "\n",
279 |     "constant = X_train.columns[~sel.get_support()]\n",
280 |     "\n",
281 |     "constant"
282 |    ]
283 |   },
284 |   {
285 |    "cell_type": "code",
286 |    "execution_count": 7,
287 |    "metadata": {},
288 |    "outputs": [
289 |     {
290 |      "data": {
291 |       "text/plain": [
292 |        "array(['x1', 'x2', 'x3', 'x4', 'x6', 'x7', 'x8'], dtype=object)"
293 |       ]
294 |      },
295 |      "execution_count": 7,
296 |      "metadata": {},
297 |      "output_type": "execute_result"
298 |     }
299 |    ],
300 |    "source": [
301 |     "sel.get_feature_names_out()"
302 |    ]
303 |   },
304 |   {
305 |    "cell_type": "code",
306 |    "execution_count": 8,
307 |    "metadata": {},
308 |    "outputs": [
309 |     {
310 |      "data": {
311 |       "text/plain": [
312 |        "((700, 7), (300, 7))"
313 |       ]
314 |      },
315 |      "execution_count": 8,
316 |      "metadata": {},
317 |      "output_type": "execute_result"
318 |     }
319 |    ],
320 |    "source": [
321 |     "# drop constant features\n",
322 |     "\n",
323 |     "X_train_t = sel.transform(X_train)\n",
324 |     "X_test_t = sel.transform(X_test)\n",
325 |     "\n",
326 |     "X_train_t.shape, X_test_t.shape"
327 |    ]
328 |   },
329 |   {
330 |    "cell_type": "code",
331 |    "execution_count": 9,
332 |    "metadata": {},
333 |    "outputs": [
334 |     {
335 |      "data": {
336 |       "text/html": [
337 |        "<div>\n",
338 |        "<style scoped>\n",
339 |        "    .dataframe tbody tr th:only-of-type {\n",
340 |        "        vertical-align: middle;\n",
341 |        "    }\n",
342 |        "\n",
343 |        "    .dataframe tbody tr th {\n",
344 |        "        vertical-align: top;\n",
345 |        "    }\n",
346 |        "\n",
347 |        "    .dataframe thead th {\n",
348 |        "        text-align: right;\n",
349 |        "    }\n",
350 |        "</style>\n",
351 |        "<table border=\"1\" class=\"dataframe\">\n",
352 |        "  <thead>\n",
353 |        "    <tr style=\"text-align: right;\">\n",
354 |        "      <th></th>\n",
355 |        "      <th>x1</th>\n",
356 |        "      <th>x2</th>\n",
357 |        "      <th>x3</th>\n",
358 |        "      <th>x4</th>\n",
359 |        "      <th>x6</th>\n",
360 |        "      <th>x7</th>\n",
361 |        "      <th>x8</th>\n",
362 |        "    </tr>\n",
363 |        "  </thead>\n",
364 |        "  <tbody>\n",
365 |        "    <tr>\n",
366 |        "      <th>0</th>\n",
367 |        "      <td>0.039801</td>\n",
368 |        "      <td>1.501392</td>\n",
369 |        "      <td>-0.189240</td>\n",
370 |        "      <td>1.546828</td>\n",
371 |        "      <td>-1.831193</td>\n",
372 |        "      <td>1.919634</td>\n",
373 |        "      <td>0.209412</td>\n",
374 |        "    </tr>\n",
375 |        "    <tr>\n",
376 |        "      <th>1</th>\n",
377 |        "      <td>-0.078494</td>\n",
378 |        "      <td>-1.536507</td>\n",
379 |        "      <td>-0.496806</td>\n",
380 |        "      <td>0.965100</td>\n",
381 |        "      <td>-0.873804</td>\n",
382 |        "      <td>-1.246872</td>\n",
383 |        "      <td>0.629114</td>\n",
384 |        "    </tr>\n",
385 |        "    <tr>\n",
386 |        "      <th>2</th>\n",
387 |        "      <td>-0.731712</td>\n",
388 |        "      <td>0.972453</td>\n",
389 |        "      <td>-0.309300</td>\n",
390 |        "      <td>-1.432922</td>\n",
391 |        "      <td>-0.419046</td>\n",
392 |        "      <td>-0.975984</td>\n",
393 |        "      <td>0.377169</td>\n",
394 |        "    </tr>\n",
395 |        "    <tr>\n",
396 |        "      <th>3</th>\n",
397 |        "      <td>-0.121187</td>\n",
398 |        "      <td>0.516685</td>\n",
399 |        "      <td>-0.800862</td>\n",
400 |        "      <td>-0.736170</td>\n",
401 |        "      <td>-1.219396</td>\n",
402 |        "      <td>-2.312341</td>\n",
403 |        "      <td>-1.027631</td>\n",
404 |        "    </tr>\n",
405 |        "    <tr>\n",
406 |        "      <th>4</th>\n",
407 |        "      <td>-2.089187</td>\n",
408 |        "      <td>0.899235</td>\n",
409 |        "      <td>-0.241111</td>\n",
410 |        "      <td>1.287536</td>\n",
411 |        "      <td>0.643273</td>\n",
412 |        "      <td>-2.310912</td>\n",
413 |        "      <td>0.085618</td>\n",
414 |        "    </tr>\n",
415 |        "  </tbody>\n",
416 |        "</table>\n",
417 |        "</div>"
418 |       ],
419 |       "text/plain": [
420 |        "         x1        x2        x3        x4        x6        x7        x8\n",
421 |        "0  0.039801  1.501392 -0.189240  1.546828 -1.831193  1.919634  0.209412\n",
422 |        "1 -0.078494 -1.536507 -0.496806  0.965100 -0.873804 -1.246872  0.629114\n",
423 |        "2 -0.731712  0.972453 -0.309300 -1.432922 -0.419046 -0.975984  0.377169\n",
424 |        "3 -0.121187  0.516685 -0.800862 -0.736170 -1.219396 -2.312341 -1.027631\n",
425 |        "4 -2.089187  0.899235 -0.241111  1.287536  0.643273 -2.310912  0.085618"
426 |       ]
427 |      },
428 |      "execution_count": 9,
429 |      "metadata": {},
430 |      "output_type": "execute_result"
431 |     }
432 |    ],
433 |    "source": [
434 |     "# sklearn returns numpy arrays. Convert to dataframe\n",
435 |     "\n",
436 |     "X_train_t = pd.DataFrame(X_train_t, columns=sel.get_feature_names_out())\n",
437 |     "X_test_t = pd.DataFrame(X_test_t, columns=sel.get_feature_names_out())\n",
438 |     "\n",
439 |     "# show result\n",
440 |     "X_train_t.head()"
441 |    ]
442 |   },
443 |   {
444 |    "cell_type": "markdown",
445 |    "metadata": {},
446 |    "source": [
447 |     "## DropConstantFeatures from Feature-engine\n",
448 |     "\n",
449 |     "Works with numerical and categorical variables."
450 |    ]
451 |   },
452 |   {
453 |    "cell_type": "code",
454 |    "execution_count": 10,
455 |    "metadata": {},
456 |    "outputs": [
457 |     {
458 |      "data": {
459 |       "text/html": [
460 |        "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>DropConstantFeatures()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DropConstantFeatures</label><div class=\"sk-toggleable__content\"><pre>DropConstantFeatures()</pre></div></div></div></div></div>"
461 |       ],
462 |       "text/plain": [
463 |        "DropConstantFeatures()"
464 |       ]
465 |      },
466 |      "execution_count": 10,
467 |      "metadata": {},
468 |      "output_type": "execute_result"
469 |     }
470 |    ],
471 |    "source": [
472 |     "# To remove constant features\n",
473 |     "sel = DropConstantFeatures(tol=1)\n",
474 |     "\n",
475 |     "# fit finds the features with only 1 value\n",
476 |     "sel.fit(X_train)  "
477 |    ]
478 |   },
479 |   {
480 |    "cell_type": "code",
481 |    "execution_count": 11,
482 |    "metadata": {},
483 |    "outputs": [
484 |     {
485 |      "data": {
486 |       "text/plain": [
487 |        "[0, 5, 9]"
488 |       ]
489 |      },
490 |      "execution_count": 11,
491 |      "metadata": {},
492 |      "output_type": "execute_result"
493 |     }
494 |    ],
495 |    "source": [
496 |     "# the constant features\n",
497 |     "\n",
498 |     "sel.features_to_drop_"
499 |    ]
500 |   },
501 |   {
502 |    "cell_type": "code",
503 |    "execution_count": 12,
504 |    "metadata": {},
505 |    "outputs": [
506 |     {
507 |      "data": {
508 |       "text/plain": [
509 |        "((700, 7), (300, 7))"
510 |       ]
511 |      },
512 |      "execution_count": 12,
513 |      "metadata": {},
514 |      "output_type": "execute_result"
515 |     }
516 |    ],
517 |    "source": [
518 |     "# drop constant features\n",
519 |     "\n",
520 |     "X_train_t = sel.transform(X_train)\n",
521 |     "X_test_t = sel.transform(X_test)\n",
522 |     "\n",
523 |     "X_train_t.shape, X_test_t.shape"
524 |    ]
525 |   },
526 |   {
527 |    "cell_type": "code",
528 |    "execution_count": 13,
529 |    "metadata": {},
530 |    "outputs": [
531 |     {
532 |      "data": {
533 |       "text/html": [
534 |        "<div>\n",
535 |        "<style scoped>\n",
536 |        "    .dataframe tbody tr th:only-of-type {\n",
537 |        "        vertical-align: middle;\n",
538 |        "    }\n",
539 |        "\n",
540 |        "    .dataframe tbody tr th {\n",
541 |        "        vertical-align: top;\n",
542 |        "    }\n",
543 |        "\n",
544 |        "    .dataframe thead th {\n",
545 |        "        text-align: right;\n",
546 |        "    }\n",
547 |        "</style>\n",
548 |        "<table border=\"1\" class=\"dataframe\">\n",
549 |        "  <thead>\n",
550 |        "    <tr style=\"text-align: right;\">\n",
551 |        "      <th></th>\n",
552 |        "      <th>1</th>\n",
553 |        "      <th>2</th>\n",
554 |        "      <th>3</th>\n",
555 |        "      <th>4</th>\n",
556 |        "      <th>6</th>\n",
557 |        "      <th>7</th>\n",
558 |        "      <th>8</th>\n",
559 |        "    </tr>\n",
560 |        "  </thead>\n",
561 |        "  <tbody>\n",
562 |        "    <tr>\n",
563 |        "      <th>105</th>\n",
564 |        "      <td>0.039801</td>\n",
565 |        "      <td>1.501392</td>\n",
566 |        "      <td>-0.189240</td>\n",
567 |        "      <td>1.546828</td>\n",
568 |        "      <td>-1.831193</td>\n",
569 |        "      <td>1.919634</td>\n",
570 |        "      <td>0.209412</td>\n",
571 |        "    </tr>\n",
572 |        "    <tr>\n",
573 |        "      <th>68</th>\n",
574 |        "      <td>-0.078494</td>\n",
575 |        "      <td>-1.536507</td>\n",
576 |        "      <td>-0.496806</td>\n",
577 |        "      <td>0.965100</td>\n",
578 |        "      <td>-0.873804</td>\n",
579 |        "      <td>-1.246872</td>\n",
580 |        "      <td>0.629114</td>\n",
581 |        "    </tr>\n",
582 |        "    <tr>\n",
583 |        "      <th>479</th>\n",
584 |        "      <td>-0.731712</td>\n",
585 |        "      <td>0.972453</td>\n",
586 |        "      <td>-0.309300</td>\n",
587 |        "      <td>-1.432922</td>\n",
588 |        "      <td>-0.419046</td>\n",
589 |        "      <td>-0.975984</td>\n",
590 |        "      <td>0.377169</td>\n",
591 |        "    </tr>\n",
592 |        "    <tr>\n",
593 |        "      <th>399</th>\n",
594 |        "      <td>-0.121187</td>\n",
595 |        "      <td>0.516685</td>\n",
596 |        "      <td>-0.800862</td>\n",
597 |        "      <td>-0.736170</td>\n",
598 |        "      <td>-1.219396</td>\n",
599 |        "      <td>-2.312341</td>\n",
600 |        "      <td>-1.027631</td>\n",
601 |        "    </tr>\n",
602 |        "    <tr>\n",
603 |        "      <th>434</th>\n",
604 |        "      <td>-2.089187</td>\n",
605 |        "      <td>0.899235</td>\n",
606 |        "      <td>-0.241111</td>\n",
607 |        "      <td>1.287536</td>\n",
608 |        "      <td>0.643273</td>\n",
609 |        "      <td>-2.310912</td>\n",
610 |        "      <td>0.085618</td>\n",
611 |        "    </tr>\n",
612 |        "  </tbody>\n",
613 |        "</table>\n",
614 |        "</div>"
615 |       ],
616 |       "text/plain": [
617 |        "            1         2         3         4         6         7         8\n",
618 |        "105  0.039801  1.501392 -0.189240  1.546828 -1.831193  1.919634  0.209412\n",
619 |        "68  -0.078494 -1.536507 -0.496806  0.965100 -0.873804 -1.246872  0.629114\n",
620 |        "479 -0.731712  0.972453 -0.309300 -1.432922 -0.419046 -0.975984  0.377169\n",
621 |        "399 -0.121187  0.516685 -0.800862 -0.736170 -1.219396 -2.312341 -1.027631\n",
622 |        "434 -2.089187  0.899235 -0.241111  1.287536  0.643273 -2.310912  0.085618"
623 |       ]
624 |      },
625 |      "execution_count": 13,
626 |      "metadata": {},
627 |      "output_type": "execute_result"
628 |     }
629 |    ],
630 |    "source": [
631 |     "# the result is already a dataframe\n",
632 |     "\n",
633 |     "X_train_t.head()"
634 |    ]
635 |   }
636 |  ],
637 |  "metadata": {
638 |   "kernelspec": {
639 |    "display_name": "fsml",
640 |    "language": "python",
641 |    "name": "fsml"
642 |   },
643 |   "language_info": {
644 |    "codemirror_mode": {
645 |     "name": "ipython",
646 |     "version": 3
647 |    },
648 |    "file_extension": ".py",
649 |    "mimetype": "text/x-python",
650 |    "name": "python",
651 |    "nbconvert_exporter": "python",
652 |    "pygments_lexer": "ipython3",
653 |    "version": "3.10.5"
654 |   },
655 |   "toc": {
656 |    "base_numbering": 1,
657 |    "nav_menu": {},
658 |    "number_sections": true,
659 |    "sideBar": true,
660 |    "skip_h1_title": false,
661 |    "title_cell": "Table of Contents",
662 |    "title_sidebar": "Contents",
663 |    "toc_cell": false,
664 |    "toc_position": {
665 |     "height": "583px",
666 |     "left": "0px",
667 |     "right": "20px",
668 |     "top": "107px",
669 |     "width": "319px"
670 |    },
671 |    "toc_section_display": "block",
672 |    "toc_window_display": true
673 |   }
674 |  },
675 |  "nbformat": 4,
676 |  "nbformat_minor": 2
677 | }
678 | 


--------------------------------------------------------------------------------
/03-Lasso.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "metadata": {},
  6 |    "source": [
  7 |     "## Lasso"
  8 |    ]
  9 |   },
 10 |   {
 11 |    "cell_type": "code",
 12 |    "execution_count": 1,
 13 |    "metadata": {},
 14 |    "outputs": [],
 15 |    "source": [
 16 |     "import numpy as np\n",
 17 |     "import pandas as pd\n",
 18 |     "from sklearn.datasets import fetch_california_housing\n",
 19 |     "from sklearn.feature_selection import SelectFromModel\n",
 20 |     "from sklearn.linear_model import Lasso\n",
 21 |     "from sklearn.model_selection import train_test_split\n",
 22 |     "from sklearn.preprocessing import StandardScaler"
 23 |    ]
 24 |   },
 25 |   {
 26 |    "cell_type": "code",
 27 |    "execution_count": 2,
 28 |    "metadata": {},
 29 |    "outputs": [],
 30 |    "source": [
 31 |     "# load the California House price data\n",
 32 |     "\n",
 33 |     "X, y = fetch_california_housing(return_X_y=True, as_frame=True)\n",
 34 |     "\n",
 35 |     "# Separate data into train and test sets\n",
 36 |     "\n",
 37 |     "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)"
 38 |    ]
 39 |   },
 40 |   {
 41 |    "cell_type": "code",
 42 |    "execution_count": 3,
 43 |    "metadata": {},
 44 |    "outputs": [
 45 |     {
 46 |      "data": {
 47 |       "text/html": [
 48 |        "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>StandardScaler()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div>"
 49 |       ],
 50 |       "text/plain": [
 51 |        "StandardScaler()"
 52 |       ]
 53 |      },
 54 |      "execution_count": 3,
 55 |      "metadata": {},
 56 |      "output_type": "execute_result"
 57 |     }
 58 |    ],
 59 |    "source": [
 60 |     "# scale the features\n",
 61 |     "\n",
 62 |     "scaler = StandardScaler()\n",
 63 |     "scaler.fit(X_train)"
 64 |    ]
 65 |   },
 66 |   {
 67 |    "cell_type": "code",
 68 |    "execution_count": 4,
 69 |    "metadata": {
 70 |     "scrolled": true
 71 |    },
 72 |    "outputs": [
 73 |     {
 74 |      "data": {
 75 |       "text/html": [
 76 |        "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SelectFromModel(estimator=Lasso(alpha=0.001, random_state=10))</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SelectFromModel</label><div class=\"sk-toggleable__content\"><pre>SelectFromModel(estimator=Lasso(alpha=0.001, random_state=10))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: Lasso</label><div class=\"sk-toggleable__content\"><pre>Lasso(alpha=0.001, random_state=10)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Lasso</label><div class=\"sk-toggleable__content\"><pre>Lasso(alpha=0.001, random_state=10)</pre></div></div></div></div></div></div></div></div></div></div>"
 77 |       ],
 78 |       "text/plain": [
 79 |        "SelectFromModel(estimator=Lasso(alpha=0.001, random_state=10))"
 80 |       ]
 81 |      },
 82 |      "execution_count": 4,
 83 |      "metadata": {},
 84 |      "output_type": "execute_result"
 85 |     }
 86 |    ],
 87 |    "source": [
 88 |     "# here, again I will train a Lasso Linear regression and select\n",
 89 |     "# the non zero features in one line.\n",
 90 |     "\n",
 91 |     "# bear in mind that the linear regression object from sklearn does\n",
 92 |     "# not allow for regularisation. So If you want to make a regularised\n",
 93 |     "# linear regression you need to import specifically \"Lasso\"\n",
 94 |     "\n",
 95 |     "sel_ = SelectFromModel(Lasso(alpha=0.001, random_state=10))\n",
 96 |     "sel_.fit(scaler.transform(X_train), y_train)"
 97 |    ]
 98 |   },
 99 |   {
100 |    "cell_type": "code",
101 |    "execution_count": 5,
102 |    "metadata": {},
103 |    "outputs": [
104 |     {
105 |      "data": {
106 |       "text/plain": [
107 |        "array([ True,  True,  True,  True,  True,  True,  True,  True])"
108 |       ]
109 |      },
110 |      "execution_count": 5,
111 |      "metadata": {},
112 |      "output_type": "execute_result"
113 |     }
114 |    ],
115 |    "source": [
116 |     "sel_.get_support()"
117 |    ]
118 |   },
119 |   {
120 |    "cell_type": "code",
121 |    "execution_count": 6,
122 |    "metadata": {},
123 |    "outputs": [
124 |     {
125 |      "name": "stdout",
126 |      "output_type": "stream",
127 |      "text": [
128 |       "total features: 8\n",
129 |       "selected features: 8\n",
130 |       "features with coefficients shrank to zero: 0\n"
131 |      ]
132 |     }
133 |    ],
134 |    "source": [
135 |     "# make a list with the selected features and print the outputs\n",
136 |     "selected_feat = X_train.columns[(sel_.get_support())]\n",
137 |     "\n",
138 |     "print('total features: {}'.format((X_train.shape[1])))\n",
139 |     "print('selected features: {}'.format(len(selected_feat)))\n",
140 |     "print('features with coefficients shrank to zero: {}'.format(\n",
141 |     "    np.sum(sel_.estimator_.coef_ == 0)))"
142 |    ]
143 |   },
144 |   {
145 |    "cell_type": "code",
146 |    "execution_count": 7,
147 |    "metadata": {},
148 |    "outputs": [
149 |     {
150 |      "data": {
151 |       "text/plain": [
152 |        "((15480, 8), (5160, 8))"
153 |       ]
154 |      },
155 |      "execution_count": 7,
156 |      "metadata": {},
157 |      "output_type": "execute_result"
158 |     }
159 |    ],
160 |    "source": [
161 |     "# we can then remove the features from the training and testing set\n",
162 |     "# like this:\n",
163 |     "\n",
164 |     "X_train_selected = sel_.transform(scaler.transform(X_train))\n",
165 |     "X_test_selected = sel_.transform(scaler.transform(X_test))\n",
166 |     "\n",
167 |     "X_train_selected.shape, X_test_selected.shape"
168 |    ]
169 |   }
170 |  ],
171 |  "metadata": {
172 |   "kernelspec": {
173 |    "display_name": "fsml",
174 |    "language": "python",
175 |    "name": "fsml"
176 |   },
177 |   "language_info": {
178 |    "codemirror_mode": {
179 |     "name": "ipython",
180 |     "version": 3
181 |    },
182 |    "file_extension": ".py",
183 |    "mimetype": "text/x-python",
184 |    "name": "python",
185 |    "nbconvert_exporter": "python",
186 |    "pygments_lexer": "ipython3",
187 |    "version": "3.10.5"
188 |   },
189 |   "toc": {
190 |    "base_numbering": 1,
191 |    "nav_menu": {},
192 |    "number_sections": true,
193 |    "sideBar": true,
194 |    "skip_h1_title": false,
195 |    "title_cell": "Table of Contents",
196 |    "title_sidebar": "Contents",
197 |    "toc_cell": false,
198 |    "toc_position": {},
199 |    "toc_section_display": "block",
200 |    "toc_window_display": true
201 |   }
202 |  },
203 |  "nbformat": 4,
204 |  "nbformat_minor": 2
205 | }
206 | 


--------------------------------------------------------------------------------
/04-Feature-shuffling.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "markdown",
  5 |    "metadata": {},
  6 |    "source": [
  7 |     "## Selection by Feature Shuffling"
  8 |    ]
  9 |   },
 10 |   {
 11 |    "cell_type": "code",
 12 |    "execution_count": 1,
 13 |    "metadata": {},
 14 |    "outputs": [],
 15 |    "source": [
 16 |     "import pandas as pd\n",
 17 |     "import matplotlib.pyplot as plt\n",
 18 |     "\n",
 19 |     "from sklearn.base import clone\n",
 20 |     "from sklearn.datasets import load_breast_cancer\n",
 21 |     "from sklearn.ensemble import RandomForestClassifier\n",
 22 |     "from sklearn.metrics import roc_auc_score\n",
 23 |     "from sklearn.model_selection import train_test_split\n",
 24 |     "\n",
 25 |     "from feature_engine.selection import SelectByShuffling"
 26 |    ]
 27 |   },
 28 |   {
 29 |    "cell_type": "code",
 30 |    "execution_count": 2,
 31 |    "metadata": {},
 32 |    "outputs": [
 33 |     {
 34 |      "data": {
 35 |       "text/html": [
 36 |        "<div>\n",
 37 |        "<style scoped>\n",
 38 |        "    .dataframe tbody tr th:only-of-type {\n",
 39 |        "        vertical-align: middle;\n",
 40 |        "    }\n",
 41 |        "\n",
 42 |        "    .dataframe tbody tr th {\n",
 43 |        "        vertical-align: top;\n",
 44 |        "    }\n",
 45 |        "\n",
 46 |        "    .dataframe thead th {\n",
 47 |        "        text-align: right;\n",
 48 |        "    }\n",
 49 |        "</style>\n",
 50 |        "<table border=\"1\" class=\"dataframe\">\n",
 51 |        "  <thead>\n",
 52 |        "    <tr style=\"text-align: right;\">\n",
 53 |        "      <th></th>\n",
 54 |        "      <th>mean radius</th>\n",
 55 |        "      <th>mean texture</th>\n",
 56 |        "      <th>mean perimeter</th>\n",
 57 |        "      <th>mean area</th>\n",
 58 |        "      <th>mean smoothness</th>\n",
 59 |        "      <th>mean compactness</th>\n",
 60 |        "      <th>mean concavity</th>\n",
 61 |        "      <th>mean concave points</th>\n",
 62 |        "      <th>mean symmetry</th>\n",
 63 |        "      <th>mean fractal dimension</th>\n",
 64 |        "      <th>...</th>\n",
 65 |        "      <th>worst radius</th>\n",
 66 |        "      <th>worst texture</th>\n",
 67 |        "      <th>worst perimeter</th>\n",
 68 |        "      <th>worst area</th>\n",
 69 |        "      <th>worst smoothness</th>\n",
 70 |        "      <th>worst compactness</th>\n",
 71 |        "      <th>worst concavity</th>\n",
 72 |        "      <th>worst concave points</th>\n",
 73 |        "      <th>worst symmetry</th>\n",
 74 |        "      <th>worst fractal dimension</th>\n",
 75 |        "    </tr>\n",
 76 |        "  </thead>\n",
 77 |        "  <tbody>\n",
 78 |        "    <tr>\n",
 79 |        "      <th>293</th>\n",
 80 |        "      <td>11.85</td>\n",
 81 |        "      <td>17.46</td>\n",
 82 |        "      <td>75.54</td>\n",
 83 |        "      <td>432.7</td>\n",
 84 |        "      <td>0.08372</td>\n",
 85 |        "      <td>0.05642</td>\n",
 86 |        "      <td>0.026880</td>\n",
 87 |        "      <td>0.022800</td>\n",
 88 |        "      <td>0.1875</td>\n",
 89 |        "      <td>0.05715</td>\n",
 90 |        "      <td>...</td>\n",
 91 |        "      <td>13.06</td>\n",
 92 |        "      <td>25.75</td>\n",
 93 |        "      <td>84.35</td>\n",
 94 |        "      <td>517.8</td>\n",
 95 |        "      <td>0.1369</td>\n",
 96 |        "      <td>0.17580</td>\n",
 97 |        "      <td>0.13160</td>\n",
 98 |        "      <td>0.09140</td>\n",
 99 |        "      <td>0.3101</td>\n",
100 |        "      <td>0.07007</td>\n",
101 |        "    </tr>\n",
102 |        "    <tr>\n",
103 |        "      <th>332</th>\n",
104 |        "      <td>11.22</td>\n",
105 |        "      <td>19.86</td>\n",
106 |        "      <td>71.94</td>\n",
107 |        "      <td>387.3</td>\n",
108 |        "      <td>0.10540</td>\n",
109 |        "      <td>0.06779</td>\n",
110 |        "      <td>0.005006</td>\n",
111 |        "      <td>0.007583</td>\n",
112 |        "      <td>0.1940</td>\n",
113 |        "      <td>0.06028</td>\n",
114 |        "      <td>...</td>\n",
115 |        "      <td>11.98</td>\n",
116 |        "      <td>25.78</td>\n",
117 |        "      <td>76.91</td>\n",
118 |        "      <td>436.1</td>\n",
119 |        "      <td>0.1424</td>\n",
120 |        "      <td>0.09669</td>\n",
121 |        "      <td>0.01335</td>\n",
122 |        "      <td>0.02022</td>\n",
123 |        "      <td>0.3292</td>\n",
124 |        "      <td>0.06522</td>\n",
125 |        "    </tr>\n",
126 |        "    <tr>\n",
127 |        "      <th>565</th>\n",
128 |        "      <td>20.13</td>\n",
129 |        "      <td>28.25</td>\n",
130 |        "      <td>131.20</td>\n",
131 |        "      <td>1261.0</td>\n",
132 |        "      <td>0.09780</td>\n",
133 |        "      <td>0.10340</td>\n",
134 |        "      <td>0.144000</td>\n",
135 |        "      <td>0.097910</td>\n",
136 |        "      <td>0.1752</td>\n",
137 |        "      <td>0.05533</td>\n",
138 |        "      <td>...</td>\n",
139 |        "      <td>23.69</td>\n",
140 |        "      <td>38.25</td>\n",
141 |        "      <td>155.00</td>\n",
142 |        "      <td>1731.0</td>\n",
143 |        "      <td>0.1166</td>\n",
144 |        "      <td>0.19220</td>\n",
145 |        "      <td>0.32150</td>\n",
146 |        "      <td>0.16280</td>\n",
147 |        "      <td>0.2572</td>\n",
148 |        "      <td>0.06637</td>\n",
149 |        "    </tr>\n",
150 |        "    <tr>\n",
151 |        "      <th>278</th>\n",
152 |        "      <td>13.59</td>\n",
153 |        "      <td>17.84</td>\n",
154 |        "      <td>86.24</td>\n",
155 |        "      <td>572.3</td>\n",
156 |        "      <td>0.07948</td>\n",
157 |        "      <td>0.04052</td>\n",
158 |        "      <td>0.019970</td>\n",
159 |        "      <td>0.012380</td>\n",
160 |        "      <td>0.1573</td>\n",
161 |        "      <td>0.05520</td>\n",
162 |        "      <td>...</td>\n",
163 |        "      <td>15.50</td>\n",
164 |        "      <td>26.10</td>\n",
165 |        "      <td>98.91</td>\n",
166 |        "      <td>739.1</td>\n",
167 |        "      <td>0.1050</td>\n",
168 |        "      <td>0.07622</td>\n",
169 |        "      <td>0.10600</td>\n",
170 |        "      <td>0.05185</td>\n",
171 |        "      <td>0.2335</td>\n",
172 |        "      <td>0.06263</td>\n",
173 |        "    </tr>\n",
174 |        "    <tr>\n",
175 |        "      <th>489</th>\n",
176 |        "      <td>16.69</td>\n",
177 |        "      <td>20.20</td>\n",
178 |        "      <td>107.10</td>\n",
179 |        "      <td>857.6</td>\n",
180 |        "      <td>0.07497</td>\n",
181 |        "      <td>0.07112</td>\n",
182 |        "      <td>0.036490</td>\n",
183 |        "      <td>0.023070</td>\n",
184 |        "      <td>0.1846</td>\n",
185 |        "      <td>0.05325</td>\n",
186 |        "      <td>...</td>\n",
187 |        "      <td>19.18</td>\n",
188 |        "      <td>26.56</td>\n",
189 |        "      <td>127.30</td>\n",
190 |        "      <td>1084.0</td>\n",
191 |        "      <td>0.1009</td>\n",
192 |        "      <td>0.29200</td>\n",
193 |        "      <td>0.24770</td>\n",
194 |        "      <td>0.08737</td>\n",
195 |        "      <td>0.4677</td>\n",
196 |        "      <td>0.07623</td>\n",
197 |        "    </tr>\n",
198 |        "  </tbody>\n",
199 |        "</table>\n",
200 |        "<p>5 rows × 30 columns</p>\n",
201 |        "</div>"
202 |       ],
203 |       "text/plain": [
204 |        "     mean radius  mean texture  mean perimeter  mean area  mean smoothness  \\\n",
205 |        "293        11.85         17.46           75.54      432.7          0.08372   \n",
206 |        "332        11.22         19.86           71.94      387.3          0.10540   \n",
207 |        "565        20.13         28.25          131.20     1261.0          0.09780   \n",
208 |        "278        13.59         17.84           86.24      572.3          0.07948   \n",
209 |        "489        16.69         20.20          107.10      857.6          0.07497   \n",
210 |        "\n",
211 |        "     mean compactness  mean concavity  mean concave points  mean symmetry  \\\n",
212 |        "293           0.05642        0.026880             0.022800         0.1875   \n",
213 |        "332           0.06779        0.005006             0.007583         0.1940   \n",
214 |        "565           0.10340        0.144000             0.097910         0.1752   \n",
215 |        "278           0.04052        0.019970             0.012380         0.1573   \n",
216 |        "489           0.07112        0.036490             0.023070         0.1846   \n",
217 |        "\n",
218 |        "     mean fractal dimension  ...  worst radius  worst texture  \\\n",
219 |        "293                 0.05715  ...         13.06          25.75   \n",
220 |        "332                 0.06028  ...         11.98          25.78   \n",
221 |        "565                 0.05533  ...         23.69          38.25   \n",
222 |        "278                 0.05520  ...         15.50          26.10   \n",
223 |        "489                 0.05325  ...         19.18          26.56   \n",
224 |        "\n",
225 |        "     worst perimeter  worst area  worst smoothness  worst compactness  \\\n",
226 |        "293            84.35       517.8            0.1369            0.17580   \n",
227 |        "332            76.91       436.1            0.1424            0.09669   \n",
228 |        "565           155.00      1731.0            0.1166            0.19220   \n",
229 |        "278            98.91       739.1            0.1050            0.07622   \n",
230 |        "489           127.30      1084.0            0.1009            0.29200   \n",
231 |        "\n",
232 |        "     worst concavity  worst concave points  worst symmetry  \\\n",
233 |        "293          0.13160               0.09140          0.3101   \n",
234 |        "332          0.01335               0.02022          0.3292   \n",
235 |        "565          0.32150               0.16280          0.2572   \n",
236 |        "278          0.10600               0.05185          0.2335   \n",
237 |        "489          0.24770               0.08737          0.4677   \n",
238 |        "\n",
239 |        "     worst fractal dimension  \n",
240 |        "293                  0.07007  \n",
241 |        "332                  0.06522  \n",
242 |        "565                  0.06637  \n",
243 |        "278                  0.06263  \n",
244 |        "489                  0.07623  \n",
245 |        "\n",
246 |        "[5 rows x 30 columns]"
247 |       ]
248 |      },
249 |      "execution_count": 2,
250 |      "metadata": {},
251 |      "output_type": "execute_result"
252 |     }
253 |    ],
254 |    "source": [
255 |     "# load dataset\n",
256 |     "\n",
257 |     "breast_cancer = load_breast_cancer()\n",
258 |     "X = pd.DataFrame(breast_cancer.data, columns=breast_cancer.feature_names)\n",
259 |     "y = breast_cancer.target\n",
260 |     "\n",
261 |     "# Separate data into train and test sets\n",
262 |     "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)\n",
263 |     "\n",
264 |     "X_train.head()"
265 |    ]
266 |   },
267 |   {
268 |    "cell_type": "code",
269 |    "execution_count": 3,
270 |    "metadata": {},
271 |    "outputs": [],
272 |    "source": [
273 |     "# the ML model for which we want to select features\n",
274 |     "\n",
275 |     "rf = RandomForestClassifier(\n",
276 |     "    n_estimators=10,\n",
277 |     "    random_state=2909,\n",
278 |     ")"
279 |    ]
280 |   },
281 |   {
282 |    "cell_type": "code",
283 |    "execution_count": 4,
284 |    "metadata": {},
285 |    "outputs": [
286 |     {
287 |      "data": {
288 |       "text/html": [
289 |        "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>SelectByShuffling(estimator=RandomForestClassifier(n_estimators=10,\n",
290 |        "                                                   random_state=2909),\n",
291 |        "                  random_state=1, threshold=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SelectByShuffling</label><div class=\"sk-toggleable__content\"><pre>SelectByShuffling(estimator=RandomForestClassifier(n_estimators=10,\n",
292 |        "                                                   random_state=2909),\n",
293 |        "                  random_state=1, threshold=0)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(n_estimators=10, random_state=2909)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestClassifier</label><div class=\"sk-toggleable__content\"><pre>RandomForestClassifier(n_estimators=10, random_state=2909)</pre></div></div></div></div></div></div></div></div></div></div>"
294 |       ],
295 |       "text/plain": [
296 |        "SelectByShuffling(estimator=RandomForestClassifier(n_estimators=10,\n",
297 |        "                                                   random_state=2909),\n",
298 |        "                  random_state=1, threshold=0)"
299 |       ]
300 |      },
301 |      "execution_count": 4,
302 |      "metadata": {},
303 |      "output_type": "execute_result"
304 |     }
305 |    ],
306 |    "source": [
307 |     "sel = SelectByShuffling(\n",
308 |     "    estimator=rf, # the ML model\n",
309 |     "    scoring='roc_auc', # the metric to evaluate\n",
310 |     "    threshold=0,# the maximum performance drop allowed to select the feature\n",
311 |     "    cv=3, # cross validation\n",
312 |     "    random_state=1 # seed\n",
313 |     ")\n",
314 |     "\n",
315 |     "sel.fit(X_train, y_train)"
316 |    ]
317 |   },
318 |   {
319 |    "cell_type": "code",
320 |    "execution_count": 5,
321 |    "metadata": {},
322 |    "outputs": [
323 |     {
324 |      "data": {
325 |       "text/plain": [
326 |        "0.9822274044237158"
327 |       ]
328 |      },
329 |      "execution_count": 5,
330 |      "metadata": {},
331 |      "output_type": "execute_result"
332 |     }
333 |    ],
334 |    "source": [
335 |     "# performance of model trained with all features\n",
336 |     "\n",
337 |     "sel.initial_model_performance_"
338 |    ]
339 |   },
340 |   {
341 |    "cell_type": "code",
342 |    "execution_count": 6,
343 |    "metadata": {},
344 |    "outputs": [
345 |     {
346 |      "data": {
347 |       "image/png": "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\n",
348 |       "text/plain": [
349 |        "<Figure size 1440x432 with 1 Axes>"
350 |       ]
351 |      },
352 |      "metadata": {
353 |       "needs_background": "light"
354 |      },
355 |      "output_type": "display_data"
356 |     }
357 |    ],
358 |    "source": [
359 |     "# Changes in performance after shuffling each feature\n",
360 |     "\n",
361 |     "pd.Series(sel.performance_drifts_).plot.bar(figsize=(20,6))\n",
362 |     "plt.title('Performance change after shuffling features')\n",
363 |     "plt.ylabel('ROC-AUC change when feature was shuffled')\n",
364 |     "plt.show()"
365 |    ]
366 |   },
367 |   {
368 |    "cell_type": "code",
369 |    "execution_count": 7,
370 |    "metadata": {},
371 |    "outputs": [
372 |     {
373 |      "data": {
374 |       "text/plain": [
375 |        "['mean smoothness',\n",
376 |        " 'mean compactness',\n",
377 |        " 'mean concavity',\n",
378 |        " 'mean symmetry',\n",
379 |        " 'mean fractal dimension',\n",
380 |        " 'radius error',\n",
381 |        " 'texture error',\n",
382 |        " 'perimeter error',\n",
383 |        " 'smoothness error',\n",
384 |        " 'compactness error',\n",
385 |        " 'concavity error',\n",
386 |        " 'concave points error',\n",
387 |        " 'symmetry error',\n",
388 |        " 'fractal dimension error',\n",
389 |        " 'worst area',\n",
390 |        " 'worst compactness',\n",
391 |        " 'worst symmetry',\n",
392 |        " 'worst fractal dimension']"
393 |       ]
394 |      },
395 |      "execution_count": 7,
396 |      "metadata": {},
397 |      "output_type": "execute_result"
398 |     }
399 |    ],
400 |    "source": [
401 |     "# the features to remove\n",
402 |     "\n",
403 |     "sel.features_to_drop_"
404 |    ]
405 |   },
406 |   {
407 |    "cell_type": "code",
408 |    "execution_count": 8,
409 |    "metadata": {},
410 |    "outputs": [
411 |     {
412 |      "data": {
413 |       "text/plain": [
414 |        "Index(['mean radius', 'mean texture', 'mean perimeter', 'mean area',\n",
415 |        "       'mean smoothness', 'mean compactness', 'mean concavity',\n",
416 |        "       'mean concave points', 'mean symmetry', 'mean fractal dimension',\n",
417 |        "       'radius error', 'texture error', 'perimeter error', 'area error',\n",
418 |        "       'smoothness error', 'compactness error', 'concavity error',\n",
419 |        "       'concave points error', 'symmetry error', 'fractal dimension error',\n",
420 |        "       'worst radius', 'worst texture', 'worst perimeter', 'worst area',\n",
421 |        "       'worst smoothness', 'worst compactness', 'worst concavity',\n",
422 |        "       'worst concave points', 'worst symmetry', 'worst fractal dimension'],\n",
423 |        "      dtype='object')"
424 |       ]
425 |      },
426 |      "execution_count": 8,
427 |      "metadata": {},
428 |      "output_type": "execute_result"
429 |     }
430 |    ],
431 |    "source": [
432 |     "# the selected features\n",
433 |     "\n",
434 |     "(pd.Series(sel.performance_drifts_)>0).index"
435 |    ]
436 |   },
437 |   {
438 |    "cell_type": "code",
439 |    "execution_count": 9,
440 |    "metadata": {},
441 |    "outputs": [
442 |     {
443 |      "data": {
444 |       "text/html": [
445 |        "<div>\n",
446 |        "<style scoped>\n",
447 |        "    .dataframe tbody tr th:only-of-type {\n",
448 |        "        vertical-align: middle;\n",
449 |        "    }\n",
450 |        "\n",
451 |        "    .dataframe tbody tr th {\n",
452 |        "        vertical-align: top;\n",
453 |        "    }\n",
454 |        "\n",
455 |        "    .dataframe thead th {\n",
456 |        "        text-align: right;\n",
457 |        "    }\n",
458 |        "</style>\n",
459 |        "<table border=\"1\" class=\"dataframe\">\n",
460 |        "  <thead>\n",
461 |        "    <tr style=\"text-align: right;\">\n",
462 |        "      <th></th>\n",
463 |        "      <th>mean radius</th>\n",
464 |        "      <th>mean texture</th>\n",
465 |        "      <th>mean perimeter</th>\n",
466 |        "      <th>mean area</th>\n",
467 |        "      <th>mean concave points</th>\n",
468 |        "      <th>area error</th>\n",
469 |        "      <th>worst radius</th>\n",
470 |        "      <th>worst texture</th>\n",
471 |        "      <th>worst perimeter</th>\n",
472 |        "      <th>worst smoothness</th>\n",
473 |        "      <th>worst concavity</th>\n",
474 |        "      <th>worst concave points</th>\n",
475 |        "    </tr>\n",
476 |        "  </thead>\n",
477 |        "  <tbody>\n",
478 |        "    <tr>\n",
479 |        "      <th>512</th>\n",
480 |        "      <td>13.40</td>\n",
481 |        "      <td>20.52</td>\n",
482 |        "      <td>88.64</td>\n",
483 |        "      <td>556.7</td>\n",
484 |        "      <td>0.08172</td>\n",
485 |        "      <td>33.67</td>\n",
486 |        "      <td>16.41</td>\n",
487 |        "      <td>29.66</td>\n",
488 |        "      <td>113.30</td>\n",
489 |        "      <td>0.15740</td>\n",
490 |        "      <td>0.51060</td>\n",
491 |        "      <td>0.20510</td>\n",
492 |        "    </tr>\n",
493 |        "    <tr>\n",
494 |        "      <th>457</th>\n",
495 |        "      <td>13.21</td>\n",
496 |        "      <td>25.25</td>\n",
497 |        "      <td>84.10</td>\n",
498 |        "      <td>537.9</td>\n",
499 |        "      <td>0.02068</td>\n",
500 |        "      <td>17.58</td>\n",
501 |        "      <td>14.35</td>\n",
502 |        "      <td>34.23</td>\n",
503 |        "      <td>91.29</td>\n",
504 |        "      <td>0.12890</td>\n",
505 |        "      <td>0.13900</td>\n",
506 |        "      <td>0.06005</td>\n",
507 |        "    </tr>\n",
508 |        "    <tr>\n",
509 |        "      <th>439</th>\n",
510 |        "      <td>14.02</td>\n",
511 |        "      <td>15.66</td>\n",
512 |        "      <td>89.59</td>\n",
513 |        "      <td>606.5</td>\n",
514 |        "      <td>0.02652</td>\n",
515 |        "      <td>19.25</td>\n",
516 |        "      <td>14.91</td>\n",
517 |        "      <td>19.31</td>\n",
518 |        "      <td>96.53</td>\n",
519 |        "      <td>0.10340</td>\n",
520 |        "      <td>0.06260</td>\n",
521 |        "      <td>0.08216</td>\n",
522 |        "    </tr>\n",
523 |        "    <tr>\n",
524 |        "      <th>298</th>\n",
525 |        "      <td>14.26</td>\n",
526 |        "      <td>18.17</td>\n",
527 |        "      <td>91.22</td>\n",
528 |        "      <td>633.1</td>\n",
529 |        "      <td>0.01374</td>\n",
530 |        "      <td>20.56</td>\n",
531 |        "      <td>16.22</td>\n",
532 |        "      <td>25.26</td>\n",
533 |        "      <td>105.80</td>\n",
534 |        "      <td>0.09445</td>\n",
535 |        "      <td>0.15650</td>\n",
536 |        "      <td>0.07530</td>\n",
537 |        "    </tr>\n",
538 |        "    <tr>\n",
539 |        "      <th>37</th>\n",
540 |        "      <td>13.03</td>\n",
541 |        "      <td>18.42</td>\n",
542 |        "      <td>82.61</td>\n",
543 |        "      <td>523.8</td>\n",
544 |        "      <td>0.02923</td>\n",
545 |        "      <td>14.16</td>\n",
546 |        "      <td>13.30</td>\n",
547 |        "      <td>22.81</td>\n",
548 |        "      <td>84.46</td>\n",
549 |        "      <td>0.09701</td>\n",
550 |        "      <td>0.04833</td>\n",
551 |        "      <td>0.05013</td>\n",
552 |        "    </tr>\n",
553 |        "  </tbody>\n",
554 |        "</table>\n",
555 |        "</div>"
556 |       ],
557 |       "text/plain": [
558 |        "     mean radius  mean texture  mean perimeter  mean area  \\\n",
559 |        "512        13.40         20.52           88.64      556.7   \n",
560 |        "457        13.21         25.25           84.10      537.9   \n",
561 |        "439        14.02         15.66           89.59      606.5   \n",
562 |        "298        14.26         18.17           91.22      633.1   \n",
563 |        "37         13.03         18.42           82.61      523.8   \n",
564 |        "\n",
565 |        "     mean concave points  area error  worst radius  worst texture  \\\n",
566 |        "512              0.08172       33.67         16.41          29.66   \n",
567 |        "457              0.02068       17.58         14.35          34.23   \n",
568 |        "439              0.02652       19.25         14.91          19.31   \n",
569 |        "298              0.01374       20.56         16.22          25.26   \n",
570 |        "37               0.02923       14.16         13.30          22.81   \n",
571 |        "\n",
572 |        "     worst perimeter  worst smoothness  worst concavity  worst concave points  \n",
573 |        "512           113.30           0.15740          0.51060               0.20510  \n",
574 |        "457            91.29           0.12890          0.13900               0.06005  \n",
575 |        "439            96.53           0.10340          0.06260               0.08216  \n",
576 |        "298           105.80           0.09445          0.15650               0.07530  \n",
577 |        "37             84.46           0.09701          0.04833               0.05013  "
578 |       ]
579 |      },
580 |      "execution_count": 9,
581 |      "metadata": {},
582 |      "output_type": "execute_result"
583 |     }
584 |    ],
585 |    "source": [
586 |     "# reduce dataset\n",
587 |     "\n",
588 |     "X_train_t = sel.transform(X_train)\n",
589 |     "X_test_t = sel.transform(X_test)\n",
590 |     "\n",
591 |     "X_test_t.head()"
592 |    ]
593 |   }
594 |  ],
595 |  "metadata": {
596 |   "kernelspec": {
597 |    "display_name": "fsml",
598 |    "language": "python",
599 |    "name": "fsml"
600 |   },
601 |   "language_info": {
602 |    "codemirror_mode": {
603 |     "name": "ipython",
604 |     "version": 3
605 |    },
606 |    "file_extension": ".py",
607 |    "mimetype": "text/x-python",
608 |    "name": "python",
609 |    "nbconvert_exporter": "python",
610 |    "pygments_lexer": "ipython3",
611 |    "version": "3.10.5"
612 |   },
613 |   "toc": {
614 |    "base_numbering": 1,
615 |    "nav_menu": {},
616 |    "number_sections": true,
617 |    "sideBar": true,
618 |    "skip_h1_title": false,
619 |    "title_cell": "Table of Contents",
620 |    "title_sidebar": "Contents",
621 |    "toc_cell": false,
622 |    "toc_position": {},
623 |    "toc_section_display": "block",
624 |    "toc_window_display": true
625 |   }
626 |  },
627 |  "nbformat": 4,
628 |  "nbformat_minor": 2
629 | }
630 | 


--------------------------------------------------------------------------------
/2022_DataTalksClub_FeatureSelection.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/solegalli/DataTalks.Club2022/a0cea4cdfccaf17fb714d0012142ce4a20a7b8c2/2022_DataTalksClub_FeatureSelection.pdf


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | BSD 3-Clause License
 2 | 
 3 | Copyright (c) - Soledad Galli
 4 | Feature Selection in Machine Learning - Book:
 5 | https://leanpub.com/feature-selection-in-machine-learning
 6 | 
 7 | 
 8 | Redistribution and use in source and binary forms, with or without
 9 | modification, are permitted provided that the following conditions are met:
10 | 
11 | 1. Redistributions of source code must retain the above copyright notice, this
12 |    list of conditions and the following disclaimer.
13 | 
14 | 2. Redistributions in binary form must reproduce the above copyright notice,
15 |    this list of conditions and the following disclaimer in the documentation
16 |    and/or other materials provided with the distribution.
17 | 
18 | 3. Neither the name of the copyright holder nor the names of its
19 |    contributors may be used to endorse or promote products derived from
20 |    this software without specific prior written permission.
21 | 
22 | 
23 | THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
24 | AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25 | IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
26 | DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
27 | FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28 | DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
29 | SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
30 | CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
31 | OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
32 | OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | ![PythonVersion](https://img.shields.io/badge/python-3.6%20|3.7%20|%203.8%20|%203.9-success)
 2 | [![License https://github.com/solegalli/DataTalks.Club2022/blob/main/LICENSE](https://img.shields.io/badge/license-BSD-success.svg)](https://github.com/solegalli/DataTalks.Club2022/blob/main/LICENSE)
 3 | [![Sponsorship https://www.trainindata.com/](https://img.shields.io/badge/Powered%20By-TrainInData-orange.svg)](https://www.trainindata.com/)
 4 | 
 5 | ## Feature Selection in Machine Learning
 6 | 
 7 | ## Links
 8 | 
 9 | - [Book](https://leanpub.com/feature-selection-in-machine-learning)
10 | - [Course](https://www.trainindata.com/p/feature-selection-for-machine-learning)
11 | 
12 | ## Talk: Contents
13 | 
14 | 1. Constant Features
15 | 2. ANOVA
16 | 3. Lasso
17 | 4. Feature Shuffling
18 | 


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