{ "cells": [ { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np\n", "from matplotlib import pyplot as plt\n", "%matplotlib inline" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Dataset \n", "We will use the \"bikes_rent.csv\" dataset, which records calendar information and weather conditions that characterize automated bike rental locations and the target: number of rentals on that day, named **cnt**. The latter we will predict them in linear regression models and evaluate models quality. Thus, we will solve the regression problem." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
seasonyrmnthholidayweekdayworkingdayweathersittempatemphumwindspeed(mph)windspeed(ms)cnt
0101060214.11084718.1812580.583310.7498824.805490985
1101000214.90259817.6869569.608716.6521137.443949801
210101118.0509249.4702543.727316.6367037.4370601349
310102118.20000010.6061059.043510.7398324.8009981562
410103119.30523711.4635043.695712.5223005.5978101600
\n", "
" ], "text/plain": [ " season yr mnth holiday weekday workingday weathersit temp \\\n", "0 1 0 1 0 6 0 2 14.110847 \n", "1 1 0 1 0 0 0 2 14.902598 \n", "2 1 0 1 0 1 1 1 8.050924 \n", "3 1 0 1 0 2 1 1 8.200000 \n", "4 1 0 1 0 3 1 1 9.305237 \n", "\n", " atemp hum windspeed(mph) windspeed(ms) cnt \n", "0 18.18125 80.5833 10.749882 4.805490 985 \n", "1 17.68695 69.6087 16.652113 7.443949 801 \n", "2 9.47025 43.7273 16.636703 7.437060 1349 \n", "3 10.60610 59.0435 10.739832 4.800998 1562 \n", "4 11.46350 43.6957 12.522300 5.597810 1600 " ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv('bikes_rent.csv')\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For each rental day, the following attributes/dataset features are known (as they were specified in the data source):\n", " - season: 1-spring, 2-summer, 3-autumn, 4-winter\n", " - yr: 0 - 2011, 1 - 2012\n", " - mnth: from 1 to 12\n", " - holiday: 0 - no holiday, 1 - there is a holiday\n", " - weekday: 0 to 6\n", " - workingday: 0 - non-working day, 1 - working day\n", " - weathersit: weather favorability rating from 1 (clear day) to 4 (rain, fog)\n", " - temp: temperature in Celsius degrees\n", " - atemp: temperature feels like, Celsius degrees\n", " - hum: humidity\n", " - windspeed(mph): wind speed in miles per hour\n", " - windspeed (ms): wind speed in meters per second\n", " - cnt: the number of rented bicycles (this is the target attribute, we will predict it thru a built model)\n", "\n", "So, we have real (*temp\tatemp, hum,\twindspeed(mph),\twindspeed(ms)*), binary (*holiday, workingday*), and nominal (ordinal) features (eg. *season: 1-spring, 2-summer, 3-autumn, 4-winter; weekday, weathersit*). All of them might be worked with as real features. \n", "\n", "One can work with nominal attributes as with real ones, because they are in an order. Let's look at the figures below, showing how the target attribute (**cnt**) depends on each of the features.\n", "\n", "#### Feature target plots" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, axes = plt.subplots(nrows=3, ncols=4, figsize=(15, 10))\n", "for idx, feature in enumerate(df.columns[:-1]):\n", " df.plot(feature, \"cnt\", subplots=True, kind=\"scatter\", ax=axes[idx // 4, idx % 4])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "One can see from the charts that there is the linear dependence of the target of some features, eg. *season, atemp*." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Features and target correlations\n", "#### Pairwise correlations between all featues except for the target\n", "Let's more strictly estimate the level of linear dependence between the features and the target variable. A good measure of the linear relationship between two vectors is the [Pearson correlation](https://en.wikipedia.org/wiki/Pearson_correlation_coefficient). In the **Pandas** it can be calculated using two dataframe methods: **corr** and **corrwith**. The **df.corr** method calculates the correlation matrix of all features from the dataframe. The **df.corrwith** method needs to supply another dataframe as an argument, and then it will calculate pairwise correlations between the features from the function argument and the original dataframe." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
seasonyrmnthholidayweekdayworkingdayweathersittempatemphumwindspeed(mph)windspeed(ms)
season1.000000-0.0018440.831440-0.010537-0.0030800.0124850.0192110.3343150.3428760.205445-0.229046-0.229046
yr-0.0018441.000000-0.0017920.007954-0.005461-0.002013-0.0487270.0476040.046106-0.110651-0.011817-0.011817
mnth0.831440-0.0017921.0000000.0191910.009509-0.0059010.0435280.2202050.2274590.222204-0.207502-0.207502
holiday-0.0105370.0079540.0191911.000000-0.101960-0.253023-0.034627-0.028556-0.032507-0.0159370.0062920.006292
weekday-0.003080-0.0054610.009509-0.1019601.0000000.0357900.031087-0.000170-0.007537-0.0522320.0142820.014282
workingday0.012485-0.002013-0.005901-0.2530230.0357901.0000000.0612000.0526600.0521820.024327-0.018796-0.018796
weathersit0.019211-0.0487270.043528-0.0346270.0310870.0612001.000000-0.120602-0.1215830.5910450.0395110.039511
temp0.3343150.0476040.220205-0.028556-0.0001700.052660-0.1206021.0000000.9917020.126963-0.157944-0.157944
atemp0.3428760.0461060.227459-0.032507-0.0075370.052182-0.1215830.9917021.0000000.139988-0.183643-0.183643
hum0.205445-0.1106510.222204-0.015937-0.0522320.0243270.5910450.1269630.1399881.000000-0.248489-0.248489
windspeed(mph)-0.229046-0.011817-0.2075020.0062920.014282-0.0187960.039511-0.157944-0.183643-0.2484891.0000001.000000
windspeed(ms)-0.229046-0.011817-0.2075020.0062920.014282-0.0187960.039511-0.157944-0.183643-0.2484891.0000001.000000
\n", "
" ], "text/plain": [ " season yr mnth holiday weekday workingday \\\n", "season 1.000000 -0.001844 0.831440 -0.010537 -0.003080 0.012485 \n", "yr -0.001844 1.000000 -0.001792 0.007954 -0.005461 -0.002013 \n", "mnth 0.831440 -0.001792 1.000000 0.019191 0.009509 -0.005901 \n", "holiday -0.010537 0.007954 0.019191 1.000000 -0.101960 -0.253023 \n", "weekday -0.003080 -0.005461 0.009509 -0.101960 1.000000 0.035790 \n", "workingday 0.012485 -0.002013 -0.005901 -0.253023 0.035790 1.000000 \n", "weathersit 0.019211 -0.048727 0.043528 -0.034627 0.031087 0.061200 \n", "temp 0.334315 0.047604 0.220205 -0.028556 -0.000170 0.052660 \n", "atemp 0.342876 0.046106 0.227459 -0.032507 -0.007537 0.052182 \n", "hum 0.205445 -0.110651 0.222204 -0.015937 -0.052232 0.024327 \n", "windspeed(mph) -0.229046 -0.011817 -0.207502 0.006292 0.014282 -0.018796 \n", "windspeed(ms) -0.229046 -0.011817 -0.207502 0.006292 0.014282 -0.018796 \n", "\n", " weathersit temp atemp hum windspeed(mph) \\\n", "season 0.019211 0.334315 0.342876 0.205445 -0.229046 \n", "yr -0.048727 0.047604 0.046106 -0.110651 -0.011817 \n", "mnth 0.043528 0.220205 0.227459 0.222204 -0.207502 \n", "holiday -0.034627 -0.028556 -0.032507 -0.015937 0.006292 \n", "weekday 0.031087 -0.000170 -0.007537 -0.052232 0.014282 \n", "workingday 0.061200 0.052660 0.052182 0.024327 -0.018796 \n", "weathersit 1.000000 -0.120602 -0.121583 0.591045 0.039511 \n", "temp -0.120602 1.000000 0.991702 0.126963 -0.157944 \n", "atemp -0.121583 0.991702 1.000000 0.139988 -0.183643 \n", "hum 0.591045 0.126963 0.139988 1.000000 -0.248489 \n", "windspeed(mph) 0.039511 -0.157944 -0.183643 -0.248489 1.000000 \n", "windspeed(ms) 0.039511 -0.157944 -0.183643 -0.248489 1.000000 \n", "\n", " windspeed(ms) \n", "season -0.229046 \n", "yr -0.011817 \n", "mnth -0.207502 \n", "holiday 0.006292 \n", "weekday 0.014282 \n", "workingday -0.018796 \n", "weathersit 0.039511 \n", "temp -0.157944 \n", "atemp -0.183643 \n", "hum -0.248489 \n", "windspeed(mph) 1.000000 \n", "windspeed(ms) 1.000000 " ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# df2 - df without the column 'cnt', the target\n", "df2 = df[df.columns[:-1]]\n", "# alternative: df2 = df[df.columns.difference(['cnt'])]\n", "# get features correlation coefficients\n", "df2.corr(method='pearson')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Pairwise correlations between featues and the target" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "season 0.406100\n", "yr 0.566710\n", "mnth 0.279977\n", "holiday -0.068348\n", "weekday 0.067443\n", "workingday 0.061156\n", "weathersit -0.297391\n", "temp 0.627494\n", "atemp 0.631066\n", "hum -0.100659\n", "windspeed(mph) -0.234545\n", "windspeed(ms) -0.234545\n", "cnt 1.000000\n", "dtype: float64" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# kitchen sink solution: pd.concat([df,df],axis=1).corr()\n", "df.corrwith(df['cnt'], axis=0, method='pearson')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The sample above shows features correlations with the target. All of them are not *null*. That means that we can solve the problem using linear methods." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The [graphs above](#Feature-target-plots) show that some of the features are similar to each other. So let's also calculate the correlations between the real features: *temp, atemp, hum, windspeed(mph), windspeed(ms) and cnt*" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
tempatemphumwindspeed(mph)windspeed(ms)cnt
temp1.0000000.9917020.126963-0.157944-0.1579440.627494
atemp0.9917021.0000000.139988-0.183643-0.1836430.631066
hum0.1269630.1399881.000000-0.248489-0.248489-0.100659
windspeed(mph)-0.157944-0.183643-0.2484891.0000001.000000-0.234545
windspeed(ms)-0.157944-0.183643-0.2484891.0000001.000000-0.234545
cnt0.6274940.631066-0.100659-0.234545-0.2345451.000000
\n", "
" ], "text/plain": [ " temp atemp hum windspeed(mph) windspeed(ms) \\\n", "temp 1.000000 0.991702 0.126963 -0.157944 -0.157944 \n", "atemp 0.991702 1.000000 0.139988 -0.183643 -0.183643 \n", "hum 0.126963 0.139988 1.000000 -0.248489 -0.248489 \n", "windspeed(mph) -0.157944 -0.183643 -0.248489 1.000000 1.000000 \n", "windspeed(ms) -0.157944 -0.183643 -0.248489 1.000000 1.000000 \n", "cnt 0.627494 0.631066 -0.100659 -0.234545 -0.234545 \n", "\n", " cnt \n", "temp 0.627494 \n", "atemp 0.631066 \n", "hum -0.100659 \n", "windspeed(mph) -0.234545 \n", "windspeed(ms) -0.234545 \n", "cnt 1.000000 " ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df3 = df[['temp', 'atemp', 'hum', 'windspeed(mph)', 'windspeed(ms)', 'cnt']]\n", "df3.corr(method='pearson')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As expected, there are ones **`1`** on the diagonals. However, there are two more pairs of strongly correlated columns in the matrix: *temp* and *atemp* (correlated by nature) and two *windspeed* features (because one is just a translation of some units into others). Later we will see that this effects negatively on the training of the linear model." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's take a look at the mean values of the features to estimate the scale of the features and the fraction of `1` for binary features." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "season 2.496580\n", "yr 0.500684\n", "mnth 6.519836\n", "holiday 0.028728\n", "weekday 2.997264\n", "workingday 0.683995\n", "weathersit 1.395349\n", "temp 20.310776\n", "atemp 23.717699\n", "hum 62.789406\n", "windspeed(mph) 12.762576\n", "windspeed(ms) 5.705220\n", "cnt 4504.348837\n", "dtype: float64" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.mean()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The features are of different scale. So for our further work, we better normalize the feature objects matrix." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Problem one: [Collinear](https://en.wikipedia.org/wiki/Multicollinearity) (highly correlated) features" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "So, in our data, some attributes duplicates another ones and there are two more very similar ones. Of course, we could remove the duplicates right away, but let's see how model would have been trained if we hadn't noticed this problem.\n", "\n", "To begin with, we will scale or standardize the features. From each feature, we will subtract its average (mean) and divide it by the standard deviation (std). This can be done using the **scale()** method.\n", "\n", "In addition we shuffle the dataset; this is required for cross-validation. " ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "from sklearn.preprocessing import scale\n", "from sklearn.utils import shuffle" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " season yr mnth holiday weekday workingday weathersit temp \\\n", "488 2 1 5 0 4 1 2 22.960000 \n", "421 1 1 2 0 0 0 1 11.445847 \n", "91 2 0 4 0 6 0 2 12.915000 \n", "300 4 0 10 0 5 1 2 13.564153 \n", "177 3 0 6 0 1 1 2 27.982500 \n", "\n", " atemp hum windspeed(mph) windspeed(ms) cnt \n", "488 26.86210 76.8333 8.957632 4.004306 6421 \n", "421 13.41540 41.0000 13.750343 6.146778 3389 \n", "91 15.78185 65.3750 13.208782 5.904686 2252 \n", "300 15.94060 58.5833 15.375093 6.873086 3747 \n", "177 31.85020 65.8333 7.208396 3.222350 4708 \n" ] } ], "source": [ "df_shuffled = shuffle(df, random_state=123)\n", "print(df_shuffled.head())\n", "X = scale(df_shuffled[df_shuffled.columns[:-1]])\n", "y = df_shuffled[\"cnt\"]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Train linear model\n", "Let's train a linear regression model on our dataset and look at the feature weights. The acquired weights are in the *coef_* parameter of a regressor class." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [], "source": [ "from sklearn.linear_model import LinearRegression" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Linear model coefficients:\n", " [ 5.70865954e+02 1.02196587e+03 -1.41301029e+02 -8.67565864e+01\n", " 1.37223034e+02 5.63936928e+01 -3.30228163e+02 3.67479222e+02\n", " 5.85554695e+02 -1.45612324e+02 1.24575770e+13 -1.24575770e+13]\n", "\n", "Coefficients and their weights:\n" ] }, { "data": { "text/plain": [ "[('season', 570.87),\n", " ('yr', 1021.97),\n", " ('mnth', -141.3),\n", " ('holiday', -86.76),\n", " ('weekday', 137.22),\n", " ('workingday', 56.39),\n", " ('weathersit', -330.23),\n", " ('temp', 367.48),\n", " ('atemp', 585.55),\n", " ('hum', -145.61),\n", " ('windspeed(mph)', 12457576985764.57),\n", " ('windspeed(ms)', -12457576985963.02)]" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn import linear_model\n", " # build a model\n", "linear_regressor = linear_model.LinearRegression() \n", " # train the model using all data\n", "linear_regressor.fit(X, y) \n", " # output coefficients\n", "print(\"Linear model coefficients:\\n\", linear_regressor.coef_ )\n", " # get the model predictions \n", " # predictions = linear_regressor.predict() \n", "print('\\nCoefficients and their weights:')\n", "[(pair[0], round(pair[1],2)) for pair in zip(df.columns, linear_regressor.coef_)]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We see that the weights for linearly dependent features are significantly greater in absolute values than for other features." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To better understand why this happened, let's recall the [analytical formula used to calculate the weights of a linear model in the least squares method](/linear-regression-stochastic-gradient-descent/#theory1):\n", "\n", "$$w = (X^TX)^{-1} X^T y$$\n", "\n", "If there are collinear (linearly dependent) columns in X, the matrix $𝑋^𝑇𝑋$ becomes degenerate one. The formula then ceases to be correct. The more dependent the features are, the smaller the determinant of this matrix is and the worse the approximation of 𝑋\\*𝑤≈𝑦 is. This situation is called the **multicollinearity problem**." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The *temp-atemp* pair had slightly less correlated variables/weights. But in practice, one should always carefully monitor the coefficients for similar attributes." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The **multicollinearity problem** solution is to regularize the linear model. The norm of weights multiplied by the regularization coefficient $\\alpha$ (L1 or L2) is added to the optimized functional $X*w$. In the first case (**L1**), the method is called Lasso, and in the second (**L2**), Ridge. More about adding regularization into Linear Regression model one can see [here](/add-regularization-in-linear-regression-model/)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We train the [Ridge](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Ridge.html) и [Lasso](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Lasso.html?highlight=lasso#sklearn.linear_model.Lasso) regressors with the default parameters and make sure that the problem with the weights is resolved." ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "from sklearn.linear_model import Lasso, Ridge" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Lasso regularization coefficients/weights:\n", " [ 5.60241616e+02 1.01946349e+03 -1.28730627e+02 -8.61527813e+01\n", " 1.37347894e+02 5.52123706e+01 -3.32369857e+02 3.76363236e+02\n", " 5.76530794e+02 -1.44129155e+02 -1.97139689e+02 -2.80504168e-08]\n", "\n", "Intercept: 4504.3488372093025\n", "\n", "Rounded coef: [560.242, 1019.463, -128.731, -86.153, 137.348, 55.212, -332.37, 376.363, 576.531, -144.129, -197.14, -0.0]\n" ] } ], "source": [ "lasso = Lasso()\n", "lasso.fit(X, y)\n", "print('Lasso regularization coefficients/weights:\\n', lasso.coef_)\n", "print('\\nIntercept:', lasso.intercept_)\n", "print( '\\nRounded coef:', list(map(lambda c: round(c, 3) , lasso.coef_) )) " ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Ridge regularization coefficients/weights:\n", " [ 563.06457225 1018.94837879 -131.87332028 -86.746098 138.00511118\n", " 55.90311038 -332.3497885 386.45788919 566.34704706 -145.0713273\n", " -99.25944108 -99.25944115]\n", "\n", "Intercept: 4504.3488372093025\n", "\n", "Rounded coef: [563.065, 1018.948, -131.873, -86.746, 138.005, 55.903, -332.35, 386.458, 566.347, -145.071, -99.259, -99.259]\n" ] } ], "source": [ "ridge = Ridge()\n", "ridge.fit(X, y)\n", "print('Ridge regularization coefficients/weights:\\n', ridge.coef_)\n", "print('\\nIntercept:', ridge.intercept_)\n", "print('\\nRounded coef:', list(map(lambda c: round(c, 3) , ridge.coef_) )) " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Problem two: Uninformative features" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In contrast to L2 regularization, L1 [resets the weights for some features](/linear-models-sklearn-linear_model-regression/#Lasso-regularizator-or-L1-for-the-Linear-Regression).\n", "\n", "#### Model coefficients as function of alpha\n", "Let's see how the weights change with an increase in the **regularization coefficient alpha**.\n", "\n", "**coefs_lasso** is the matrix of weights of size (number of regressors) x (number of features)." ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [], "source": [ "alphas = np.arange(1, 500, 50)\n", "coefs_lasso = np.zeros((alphas.shape[0], X.shape[1])) " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For each coefficient value from **alphas** set we train the Lasso regressor and write the weights into the corresponding rows of the **coefs_lasso** matrix. Then we train the Ridge regressor and write the weights into **coefs_ridge** matrix." ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Matrix `coefs_lasso` size: (10, 12) \n", "Alpha values: [ 1 51 101 151 201 251 301 351 401 451]\n", "Lasso coefficients, head:\n", " [[ 5.60241616e+02 1.01946349e+03 -1.28730627e+02 -8.61527813e+01\n", " 1.37347894e+02 5.52123706e+01 -3.32369857e+02 3.76363236e+02\n", " 5.76530794e+02 -1.44129155e+02 -1.97139689e+02 -2.80504168e-08]\n", " [ 4.10969632e+02 9.77019409e+02 -0.00000000e+00 -5.34489688e+01\n", " 9.19434374e+01 1.75372118e+01 -3.18125568e+02 3.22829934e+02\n", " 6.10031512e+02 -9.10689615e+01 -1.45066095e+02 -2.29877063e-08]\n", " [ 3.70077089e+02 9.35945490e+02 0.00000000e+00 -1.21619360e+01\n", " 4.88886342e+01 0.00000000e+00 -3.08805664e+02 2.69417263e+02\n", " 6.32502623e+02 -2.75042876e+01 -9.37749037e+01 -2.41652170e-08]\n", " [ 3.32835717e+02 8.91870058e+02 0.00000000e+00 -0.00000000e+00\n", " 0.00000000e+00 0.00000000e+00 -2.79616688e+02 2.11052030e+02\n", " 6.62920880e+02 -0.00000000e+00 -5.01551472e+01 -2.62761604e-08]\n", " [ 2.98134448e+02 8.45652857e+02 0.00000000e+00 -0.00000000e+00\n", " 0.00000000e+00 0.00000000e+00 -2.35571345e+02 1.24144807e+02\n", " 7.25379483e+02 -0.00000000e+00 -1.26461769e+01 -2.78781252e-08]]\n", "\n", "Rounded Lasso coefficients:\n", " [[560.242, 1019.463, -128.731, -86.153, 137.348, 55.212, -332.37, 376.363, 576.531, -144.129, -197.14, -0.0], [410.97, 977.019, -0.0, -53.449, 91.943, 17.537, -318.126, 322.83, 610.032, -91.069, -145.066, -0.0], [370.077, 935.945, 0.0, -12.162, 48.889, 0.0, -308.806, 269.417, 632.503, -27.504, -93.775, -0.0], [332.836, 891.87, 0.0, -0.0, 0.0, 0.0, -279.617, 211.052, 662.921, -0.0, -50.155, -0.0], [298.134, 845.653, 0.0, -0.0, 0.0, 0.0, -235.571, 124.145, 725.379, -0.0, -12.646, -0.0], [258.927, 799.237, 0.0, -0.0, 0.0, 0.0, -190.822, 72.076, 750.363, -0.0, -0.0, -0.0], [217.428, 752.721, 0.0, -0.0, 0.0, 0.0, -145.713, 37.715, 756.297, -0.0, -0.0, -0.0], [175.93, 706.204, 0.0, -0.0, 0.0, 0.0, -100.605, 3.661, 761.926, -0.0, -0.0, -0.0], [134.628, 659.633, 0.0, -0.0, 0.0, 0.0, -55.511, 0.0, 737.348, -0.0, -0.0, -0.0], [93.352, 613.055, 0.0, -0.0, 0.0, 0.0, -10.419, 0.0, 709.131, -0.0, -0.0, -0.0]]\n" ] } ], "source": [ "print('Matrix `coefs_lasso` size:', (alphas.shape[0], X.shape[1]), \"\\nAlpha values:\", alphas )\n", "for i, al in enumerate(alphas):\n", " lasso = Lasso(alpha = al)\n", " lasso.fit(X, y)\n", " for j, coef in enumerate(lasso.coef_):\n", " coefs_lasso[i][j]=coef\n", " #print(list(map(lambda c: round(c, 3) , coefs_lasso[i]) ))\n", "# list(map(lambda c: round(c, 3) , coefs_lasso) ))\n", "print('Lasso coefficients, head:\\n', coefs_lasso[:5])\n", "print('\\nRounded Lasso coefficients:\\n', list(map(lambda i: list(map(lambda j: round(j, 3), i)), coefs_lasso)))\n", " " ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[563.065, 1018.948, -131.873, -86.746, 138.005, 55.903, -332.35, 386.458, 566.347, -145.071, -99.259, -99.259]\n", "[461.179, 954.308, -41.565, -84.913, 126.604, 54.252, -313.275, 458.901, 481.444, -151.291, -101.627, -101.627]\n", "[403.977, 898.084, 5.674, -81.911, 117.941, 52.728, -298.409, 455.29, 467.431, -152.686, -102.102, -102.102]\n", "[366.604, 848.463, 34.027, -78.772, 110.68, 51.257, -286.125, 447.48, 455.754, -151.483, -102.005, -102.005]\n", "[339.745, 804.251, 52.49, -75.717, 104.403, 49.842, -275.486, 438.51, 444.764, -148.944, -101.586, -101.586]\n", "[319.159, 764.561, 65.152, -72.82, 98.879, 48.485, -266.003, 429.214, 434.235, -145.698, -100.965, -100.965]\n", "[302.636, 728.709, 74.138, -70.099, 93.956, 47.185, -257.391, 419.926, 424.122, -142.089, -100.209, -100.209]\n", "[288.913, 696.146, 80.66, -67.555, 89.529, 45.943, -249.471, 410.8, 414.409, -138.317, -99.361, -99.361]\n", "[277.213, 666.43, 85.459, -65.179, 85.519, 44.757, -242.124, 401.912, 405.086, -134.499, -98.449, -98.449]\n", "[267.031, 639.195, 89.012, -62.959, 81.865, 43.625, -235.264, 393.298, 396.138, -130.71, -97.493, -97.493]\n" ] } ], "source": [ "coefs_ridge = np.zeros((alphas.shape[0], X.shape[1]))\n", "for i, al in enumerate(alphas):\n", " ridge = Ridge(alpha = al)\n", " ridge.fit(X, y)\n", " for j, coef in enumerate(ridge.coef_):\n", " coefs_ridge[i][j]=coef\n", " print(list(map(lambda c: round(c, 3) , coefs_ridge[i]) ))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### The weights dependence on alpha dynamics\n", "We visualize the dynamics of the weights as the regularization parameter $\\alpha$ increases." ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 1.0, 'Ridge')" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(10, 8))\n", "for coef, feature in zip(coefs_lasso.T, df.columns):\n", " plt.plot(alphas, coef, label=feature, color=np.random.rand(3))\n", "plt.legend(loc=\"upper right\", bbox_to_anchor=(1.4, 0.95))\n", "plt.xlabel(\"Alpha\")\n", "plt.ylabel(\"Feature weight\")\n", "plt.title(\"Lasso\")\n", "\n", "plt.figure(figsize=(10, 8))\n", "for coef, feature in zip(coefs_ridge.T, df.columns):\n", " plt.plot(alphas, coef, label=feature, color=np.random.rand(3))\n", "plt.legend(loc=\"upper right\", bbox_to_anchor=(1.4, 0.95))\n", "plt.xlabel(\"Alpha\")\n", "plt.ylabel(\"Feature weight\")\n", "plt.title(\"Ridge\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Subconclusions\n", "\n", "We may do the following conclusions of the above figures:\n", " 1. Lasso regularizator reduces weights more aggressively than Ridge at the same alpha. \n", " 2. The regression model weights with the Lasso regularizator will tend/pursure to zero. The higher the alpha, the lower the complexity of the model. Then, for very large values of alpha, it is optimal to simply bring all the weights to zero.\n", " 3. It is assumed that the regularizator excludes a feature if the feature coefficient is less than 0.001 (`<1e-3`). At Lasso regularizator with `alpha = 1`, the weights for both \"windspeeds\" are not zero. But for `alpha >= 50`, the weights for both \"windspeeds\" are greatly reduced. As for the Ridge (L2) regularizator, the weights for collinear features depend little on alpha. Ridge does not exclude features when alpha increasing.\n", " 4. Lasso regularizator is more fitting for selecting uninformative features. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### We will work with Lasso (L1) regularizator to optimize regression model with regularization\n", "\n", "We might see that when alpha changes, the model selects the feature coefficients differently. We need to choose the best alpha then." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We first choose a quality metric. We will take the optimized functional of the least squares method, i.e. Mean Square Error (MSE), as a metric. Now, how to calculate MSE for different alpha?\n", "#### Cross-validation\n", "\n", " - We can't choose alpha by the MSE value on the whole training sample\n", " - We neither can choose a single split of the sample into training and test sets (\"holdout\"), since we will tune the model in to specific \"new\" data, namely the test set.\n", " - The **best approach is cross-validation**. We divide the sample into `K` parts, or blocks, and each time take one of them as a test, and make a training sample from the remaining blocks. We will make several partitions of the sample, try different alpha values on each one, and then average the MSE. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The cross-validation for regression in **sklearn** library is quite simple: there is a special regressor for this, namely [__LassoCV__](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LassoCV.html). It takes as input a list of alphas and calculates the MSE for each of them on cross-validation. After training (parameter `cv=3`), the regressor will contain the variable __mse\\_path\\___, a matrix of size `len(alpha) x K`, where `K = 3` (the number of blocks in cross-validation), containing the MSE values on the test for the corresponding runs. In addition, the variable __alpha\\___ will store the selected (optimal) value of the regularization parameter, and in __coef\\___, will have the trained weights corresponding to this __alpha___.\n", "\n", "Note that the regressor can change the order in which it iterates through the alphas values. In order to map it to the MSE matrix, it is better to use the regressor variable __alphas___." ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [], "source": [ "from sklearn.linear_model import LassoCV" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Regularizator score (coefficient of determination R^2 of the prediction): 0.800081404381094\n", "\n", "Regularizator MSE matrix head:\n", " [[863936.50981215 826364.11936907 862993.29751897]\n", " [860479.31511364 821110.1817776 853075.13780626]\n", " [857344.83606082 816153.27782428 843628.81286099]\n", " [854526.73639431 811496.34805693 834654.45357263]\n", " [852024.62341384 807139.39657173 826152.16399016]]\n", "\n", "Regularizator's `Alpha`: 6\n", "\n", "Regularizator coefficients:\n" ] }, { "data": { "text/plain": [ "[('season', 532.019),\n", " ('yr', 1015.0602),\n", " ('mnth', -100.0395),\n", " ('holiday', -83.294),\n", " ('weekday', 132.5045),\n", " ('workingday', 51.5571),\n", " ('weathersit', -330.5599),\n", " ('temp', 370.6799),\n", " ('atemp', 581.3969),\n", " ('hum', -140.0074),\n", " ('windspeed(mph)', -191.7714),\n", " ('windspeed(ms)', -0.0)]" ] }, "execution_count": 23, "metadata": {}, "output_type": "execute_result" } ], "source": [ "alphas = np.arange(1, 100, 5)\n", "reg = LassoCV(cv=3, random_state=0, alphas = alphas )\n", "reg.fit(X, y)\n", "print('Regularizator score (coefficient of determination R^2 of the prediction):', reg.score(X, y))\n", "print('\\nRegularizator MSE matrix head:\\n', reg.mse_path_[:5])\n", "print('\\nRegularizator\\'s `Alpha`:', reg.alpha_)\n", "print('\\nRegularizator coefficients:') \n", "[(pair[0], round(pair[1], 4)) for pair in zip(df.columns[:-1], reg.coef_)]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "#### Plotting the averaged MSE as a function of alpha." ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0.5, 1.0, 'Lasso regularizator MSE')" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "mean_mse_arr = np.array(reg.mse_path_).mean(axis=1)\n", "#print(mean_mse_arr)\n", "plt.figure(figsize=(10, 8)) \n", "plt.plot(reg.alphas_, mean_mse_arr, label=feature ) \n", "plt.xlabel(\"Alpha\")\n", "plt.ylabel(\"Mean MSE by fold\")\n", "plt.title(\"Lasso regularizator MSE\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We have chosen a certain regularization parameter (alpha). Let's look at what kind of alpha we would choose if we divided the dataset only once into training and test sets. We'll consider the MSE graphs corresponding to single blocks (fold) of the dataset." ] }, { "cell_type": "code", "execution_count": 31, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Minimum MSE for each column: [843336.18149882 772598.49562777 745668.60596081] \n", "\n", "Values of alpha, corresponding to MSE at each column:\n", "column 0, alpha = 41\n", "column 1, alpha = 6\n", "column 2, alpha = 1\n" ] } ], "source": [ "mse_columns = np.array(reg.mse_path_).transpose()\n", "mins = np.amin(mse_columns, axis=1)\n", "print('Minimum MSE for each column:', mins , '\\n' ) \n", "print ('Values of alpha, corresponding to MSE at each column:')\n", "\n", "for i in range(3):\n", " for j in range(len(reg.mse_path_)):\n", " if reg.mse_path_[j, i] == mins[i]:\n", " alpha = reg.alphas_[j]\n", " print ('column {0}, alpha = {1}'.format(i, alpha) )" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Mean MSE for column 0: 848818.027469035\n", "Mean MSE for column 1: 797903.087624002\n", "Mean MSE for column 2: 794699.8496187156\n" ] }, { "data": { "text/plain": [ "Text(0.5, 1.0, 'Lasso regularizator MSE')" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "for i, column in enumerate(mse_columns):\n", " print('Mean MSE for column {}:'.format(i) , column.mean() )\n", "plt.figure(figsize=(10, 8)) \n", "plt.plot(reg.alphas_, mse_columns[0], label=feature ) \n", "plt.xlabel(\"Alpha\")\n", "plt.ylabel(\"MSE in the 1-st fold, index 0\")\n", "plt.title(\"Lasso regularizator MSE\")" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, axs = plt.subplots(len(mse_columns))\n", "fig.set_size_inches(12,12)\n", "fig.suptitle('MSE in each of 3 Folds')\n", "for i, col in enumerate(mse_columns):\n", " axs[i].plot(reg.alphas_, col) \n", " axs[i].set_xlabel('Alpha')\n", " axs[i].set_ylabel('MSE, fold {}'.format(i))\n", " " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "On each partition (fold), the optimal value of **alpha** is different. It might correspond to a large MSE on the other partitions. It turns out that we are tuning regularization parameter **alpha** in to specific training and control sets. When choosing **alpha** on cross-validation, we choose something \"average\" that will give an acceptable metric value on different dataset partitions." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Sub-conclusion on the Lasso regularization\n", " 1. In the last trained model, there are 4 features with the **highest (positive) coefficients**: *'season', 'yr', 'temp', 'atemp'*. \n", " 2. We may look at the [visualizations of the **cnt** dependencies](#Feature-target-plots) on these features that we drew in the \"Dataset\" block. There we may see increasing linear dependence of **cnt** on these features. It is logical to say (from common sense) that the greater the value of these feature weights, the more people will want to take bicycles. \n", " 3. We select the features with the largest **modulo negative coefficients**: *'weathersit', ' windspeed(mph)', 'hum'*.\n", " 4. The graphs show a decreasing linear relationship with the target, `cnt`. It is logical to say that the greater the magnitude of these feature weights, the fewer people will want to take bicycles.\n", " 5. The feature *windspeed (ms)* has coefficient close to zero (less than `1e-3`). The linear model excluded it since it is a correlated (collinear) feature with *windspeed(mph)* from the model. One may see it at the graphs there. It is true that this feature does not affect the demand for bicycles since *windspeed (ms)* has already influenced it." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Conclusion\n", "In the post we've built a linear model and evaluated the adequacy of the model. We've learned how to select features, and how to correctly select the regularization coefficient `alpha` (without tuning in the model to any particular piece of data).\n", "\n", "When using the cross-validation we selected only a small number of folds/partitions (`3`). It's because for each valid combination of them, we have to train the model several times (for different **alpha**) that is a time-consuming process esp. for large amounts of data." ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "anaconda-cloud": {}, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.5" } }, "nbformat": 4, "nbformat_minor": 1 }