{ "cells": [ { "cell_type": "markdown", "id": "5f22a293-26ad-49a6-b617-53ab617dc5ae", "metadata": { "tags": [] }, "source": [ "# Preparing data\n", "1. Loading data\n", "1. Impute missing values\n", "2. Rescaling/transforming data\n", "3. Feature engineering" ] }, { "cell_type": "markdown", "id": "4bd639f9-0a7f-48ca-a372-a19d614134d7", "metadata": { "tags": [] }, "source": [ "## Loading Data" ] }, { "cell_type": "markdown", "id": "3f93b3d2-afba-4300-a176-ebdbf51185a9", "metadata": { "tags": [] }, "source": [ "### Sklearn datasets\n", "\n", "The `sklearn.datasets` package embeds some small sample [datasets](https://scikit-learn.org/stable/datasets.html)\n", "\n", "For each dataset, there are 4 varibles:\n", "\n", "- **data**: numpy array of predictors/X\n", "- **target**: numpy array of predictant/target/y\n", "- **feature_names**: names of all predictors in X\n", "- **target_names**: names of all predictand in y\n", "\n", "For example:" ] }, { "cell_type": "code", "execution_count": 1, "id": "12f8bf98-7cf9-4d9c-8a9f-7f38847b8f78", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[5.1 3.5 1.4 0.2]\n", " [4.9 3. 1.4 0.2]\n", " [4.7 3.2 1.3 0.2]\n", " [4.6 3.1 1.5 0.2]\n", " [5. 3.6 1.4 0.2]]\n", "[0 0 0 0 0]\n", "['sepal length (cm)', 'sepal width (cm)', 'petal length (cm)', 'petal width (cm)']\n", "['setosa' 'versicolor' 'virginica']\n" ] } ], "source": [ "from sklearn.datasets import load_iris\n", "data = load_iris()\n", "print(data.data[:5])\n", "print(data.target[:5])\n", "print(data.feature_names)\n", "print(data.target_names)" ] }, { "cell_type": "markdown", "id": "dcef81f4-e78d-4eb0-b181-ee43ed26bd1c", "metadata": {}, "source": [ "### Pandas" ] }, { "cell_type": "code", "execution_count": 2, "id": "6ed874b4-2c50-4f10-9d64-8b5a13fe1a90", "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", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
OzoneSolar.RWindTempMonthDay
041.0190.07.46751
136.0118.08.07252
212.0149.012.67453
318.0313.011.56254
4NaNNaN14.35655
.....................
14830.0193.06.970926
149NaN145.013.277927
15014.0191.014.375928
15118.0131.08.076929
15220.0223.011.568930
\n", "

153 rows × 6 columns

\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.0 190.0 7.4 67 5 1\n", "1 36.0 118.0 8.0 72 5 2\n", "2 12.0 149.0 12.6 74 5 3\n", "3 18.0 313.0 11.5 62 5 4\n", "4 NaN NaN 14.3 56 5 5\n", ".. ... ... ... ... ... ...\n", "148 30.0 193.0 6.9 70 9 26\n", "149 NaN 145.0 13.2 77 9 27\n", "150 14.0 191.0 14.3 75 9 28\n", "151 18.0 131.0 8.0 76 9 29\n", "152 20.0 223.0 11.5 68 9 30\n", "\n", "[153 rows x 6 columns]" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "data_df = pd.DataFrame(pd.read_csv('/zfs/citi/workshop_data/python_ml/r_airquality.csv'))\n", "data_df" ] }, { "cell_type": "markdown", "id": "80d1f680-443f-4335-9a48-20b140fa1689", "metadata": { "tags": [] }, "source": [ "## Impute missing values" ] }, { "cell_type": "markdown", "id": "13a9c9ee-8d37-490a-8b5c-cd734c9f4c73", "metadata": {}, "source": [ "- There are several ways to treat the missing values:\n", "- Method 1: remove all missing `NA` values" ] }, { "cell_type": "code", "execution_count": 3, "id": "25666ac6-8027-4f76-b9eb-f1bf241577cc", "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", "
OzoneSolar.RWindTempMonthDay
041.0190.07.46751
136.0118.08.07252
212.0149.012.67453
318.0313.011.56254
623.0299.08.66557
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.0 190.0 7.4 67 5 1\n", "1 36.0 118.0 8.0 72 5 2\n", "2 12.0 149.0 12.6 74 5 3\n", "3 18.0 313.0 11.5 62 5 4\n", "6 23.0 299.0 8.6 65 5 7" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data1 = data_df.dropna()\n", "data1.head()" ] }, { "cell_type": "markdown", "id": "57dd3219-9704-482b-804f-06beaa9402d8", "metadata": {}, "source": [ "- Method 2: Set `NA` to mean value " ] }, { "cell_type": "code", "execution_count": 4, "id": "4a448992-d682-40da-90e2-e5496310350f", "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", "
OzoneSolar.RWindTempMonthDay
041.00000190.0000007.46751
136.00000118.0000008.07252
212.00000149.00000012.67453
318.00000313.00000011.56254
442.12931185.93150714.35655
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.00000 190.000000 7.4 67 5 1\n", "1 36.00000 118.000000 8.0 72 5 2\n", "2 12.00000 149.000000 12.6 74 5 3\n", "3 18.00000 313.000000 11.5 62 5 4\n", "4 42.12931 185.931507 14.3 56 5 5" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data2 = data_df.copy()\n", "data2.fillna(data2.mean(), inplace=True)\n", "data2.head()" ] }, { "cell_type": "markdown", "id": "bb5660e0-5e97-4c09-8677-49fef7477bca", "metadata": {}, "source": [ "- Method 3: Use `Impute` to handle missing values\n", "\n", "In statistics, imputation is the process of replacing missing data with substituted values. \n", "Because missing data can create problems for analyzing data, imputation is seen as a way to \n", "avoid pitfalls involved with listwise deletion of cases that have missing values. That is to \n", "say, when one or more values are missing for a case, most statistical packages default to discarding \n", "any case that has a missing value, which may introduce bias or affect the representativeness of the \n", "results. Imputation preserves all cases by replacing missing data with an estimated value based on \n", "other available information. Once all missing values have been imputed, the data set can then be \n", "analysed using standard techniques for complete data. There have been many theories embraced by \n", "scientists to account for missing data but the majority of them introduce bias. A few of the well \n", "known attempts to deal with missing data include: \n", "- hot deck and cold deck imputation; \n", "- listwise and pairwise deletion; \n", "- mean imputation; \n", "- non-negative matrix factorization; \n", "- regression imputation; \n", "- last observation carried forward; \n", "- stochastic imputation; \n", "- and multiple imputation." ] }, { "cell_type": "code", "execution_count": 5, "id": "0925abd1-0d07-4696-acb7-c3f6873c79f1", "metadata": { "tags": [] }, "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", "
OzoneSolar.RWindTempMonthDay
041.00000190.0000007.467.05.01.0
136.00000118.0000008.072.05.02.0
212.00000149.00000012.674.05.03.0
318.00000313.00000011.562.05.04.0
442.12931185.93150714.356.05.05.0
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.00000 190.000000 7.4 67.0 5.0 1.0\n", "1 36.00000 118.000000 8.0 72.0 5.0 2.0\n", "2 12.00000 149.000000 12.6 74.0 5.0 3.0\n", "3 18.00000 313.000000 11.5 62.0 5.0 4.0\n", "4 42.12931 185.931507 14.3 56.0 5.0 5.0" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import numpy as np\n", "from sklearn.impute import SimpleImputer\n", "imputer = SimpleImputer(missing_values=np.nan, strategy='mean')\n", "data3 = pd.DataFrame(imputer.fit_transform(data_df))\n", "data3.columns = data_df.columns\n", "data3.head()" ] }, { "cell_type": "markdown", "id": "50a49404-f9f2-4ca2-994d-5142821b4352", "metadata": {}, "source": [ "**Note:**\n", "SimpleImputer converts missing values to **mean, median, most_frequent and constant**.\n", "\n", "---\n", "\n", "**Question**: How do we decide the right imputation strategy? \n", "\n", "---" ] }, { "cell_type": "code", "execution_count": 6, "id": "7f576bcd-bc31-4abd-99f6-cb9447e9066c", "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", "
OzoneSolar.RWindTempMonthDay
041.0190.07.467.05.01.0
136.0118.08.072.05.02.0
212.0149.012.674.05.03.0
318.0313.011.562.05.04.0
431.5205.014.356.05.05.0
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.0 190.0 7.4 67.0 5.0 1.0\n", "1 36.0 118.0 8.0 72.0 5.0 2.0\n", "2 12.0 149.0 12.6 74.0 5.0 3.0\n", "3 18.0 313.0 11.5 62.0 5.0 4.0\n", "4 31.5 205.0 14.3 56.0 5.0 5.0" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "imputer = SimpleImputer(missing_values=np.nan, strategy='median')\n", "data4 = pd.DataFrame(imputer.fit_transform(data_df))\n", "data4.columns = data_df.columns\n", "data4.head()" ] }, { "cell_type": "markdown", "id": "81ba2299-5711-4dbd-91be-181bd032ffc8", "metadata": {}, "source": [ "`knnImpute` can also be used to fill in missing value" ] }, { "cell_type": "code", "execution_count": 7, "id": "1661af82-9e46-47a3-a638-9f9e1c7024b2", "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", "
012345
041.0190.07.467.05.01.0
136.0118.08.072.05.02.0
212.0149.012.674.05.03.0
318.0313.011.562.05.04.0
418.5206.014.356.05.05.0
\n", "
" ], "text/plain": [ " 0 1 2 3 4 5\n", "0 41.0 190.0 7.4 67.0 5.0 1.0\n", "1 36.0 118.0 8.0 72.0 5.0 2.0\n", "2 12.0 149.0 12.6 74.0 5.0 3.0\n", "3 18.0 313.0 11.5 62.0 5.0 4.0\n", "4 18.5 206.0 14.3 56.0 5.0 5.0" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.impute import KNNImputer\n", "imputer = KNNImputer(n_neighbors=2, weights=\"uniform\")\n", "data_knnimpute = pd.DataFrame(imputer.fit_transform(data_df))\n", "data_knnimpute.head()" ] }, { "cell_type": "markdown", "id": "75ab8133-4b41-4484-b350-78636e43fd03", "metadata": { "tags": [] }, "source": [ "- In addition to KNNImputer, there are **IterativeImputer** (Multivariate imputer that estimates each feature from all the others) and **MissingIndicator**(Binary indicators for missing values)\n", "- More information on sklearn.impute can be found [here](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.impute)" ] }, { "cell_type": "markdown", "id": "656ecb3b-5945-4f96-8db3-1e77350d3beb", "metadata": { "tags": [] }, "source": [ "## Standardizing data\n", "![image](https://user-images.githubusercontent.com/43855029/114231774-df6ba180-9948-11eb-9c61-3d2e0d3df889.png)\n", "\n", "- Standardization comes into picture when features of input data set have large differences between their ranges, or simply when they are measured in different measurement units for example: rainfall (0-1000mm), temperature (-10 to 40oC), humidity (0-100%), etc.\n", "- Standardition Convert all independent variables into the same scale (mean=0, std=1) \n", "- These differences in the ranges of initial features causes trouble to many machine learning models. For example, for the models that are based on distance computation, if one of the features has a broad range of values, the distance will be governed by this particular feature.\n", "- The example below use data from above:" ] }, { "cell_type": "code", "execution_count": 8, "id": "3b589291-4aa4-4a6a-b205-367573683481", "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", "
OzoneSolar.RWindTempMonthDay
0-0.0394870.046406-0.728332-1.153490-1.411916-1.675504
1-0.214316-0.774834-0.557464-0.623508-1.411916-1.562324
2-1.053493-0.4212450.752529-0.411515-1.411916-1.449144
3-0.8436981.4493570.439270-1.683472-1.411916-1.335965
40.0000000.0000001.236657-2.319450-1.411916-1.222785
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 -0.039487 0.046406 -0.728332 -1.153490 -1.411916 -1.675504\n", "1 -0.214316 -0.774834 -0.557464 -0.623508 -1.411916 -1.562324\n", "2 -1.053493 -0.421245 0.752529 -0.411515 -1.411916 -1.449144\n", "3 -0.843698 1.449357 0.439270 -1.683472 -1.411916 -1.335965\n", "4 0.000000 0.000000 1.236657 -2.319450 -1.411916 -1.222785" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.preprocessing import scale\n", "data_std = pd.DataFrame(scale(data3, axis=0, with_mean=True, with_std=True))\n", "data_std.columns = data3.columns\n", "data_std.head()" ] }, { "cell_type": "code", "execution_count": 9, "id": "de5dca29-ac22-4802-bc53-00db743740d3", "metadata": {}, "outputs": [ { "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": [ "import matplotlib.pyplot as plt\n", "plt.hist(data_df['Temp'], bins=20)\n", "plt.title(\"Original\")\n", "plt.show()\n", "\n", "plt.hist(data_std['Temp'], bins=20)\n", "plt.title(\"Standardized\")\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "e0515d17-624c-49d9-a82a-56539bae83a4", "metadata": {}, "source": [ "#### 2.3.2.2 Using scaling with predefine range\n", "Transform features by scaling each feature to a given range.\n", "This estimator scales and translates each feature individually such that it is in the given range on the training set, e.g. between zero and one.\n", "Formulation for this is:\n", "\n", "```\n", "X_std = (X - X.min(axis=0)) / (X.max(axis=0) - X.min(axis=0))\n", "X_scaled = X_std * (max - min) + min\n", "```" ] }, { "cell_type": "code", "execution_count": 10, "id": "586d9d86-5c5b-493a-b434-f0374ca40a87", "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", "
OzoneSolar.RWindTempMonthDay
00.2395210.5596330.3000000.2682930.00.000000
10.2095810.3394500.3315790.3902440.00.033333
20.0658680.4342510.5736840.4390240.00.066667
30.1017960.9357800.5157890.1463410.00.100000
40.2462830.5471910.6631580.0000000.00.133333
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 0.239521 0.559633 0.300000 0.268293 0.0 0.000000\n", "1 0.209581 0.339450 0.331579 0.390244 0.0 0.033333\n", "2 0.065868 0.434251 0.573684 0.439024 0.0 0.066667\n", "3 0.101796 0.935780 0.515789 0.146341 0.0 0.100000\n", "4 0.246283 0.547191 0.663158 0.000000 0.0 0.133333" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.preprocessing import MinMaxScaler\n", "scaler = MinMaxScaler()\n", "data_scaler = pd.DataFrame(scaler.fit_transform(data3), columns = data3.columns)\n", "data_scaler.head()" ] }, { "cell_type": "code", "execution_count": 11, "id": "d0f5b9ca-ce6d-4cf9-8944-07456b5ff21f", "metadata": {}, "outputs": [ { "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.hist(data_df['Temp'], bins=20)\n", "plt.title(\"Original\")\n", "plt.show()\n", "\n", "plt.hist(data_scaler['Temp'], bins=20)\n", "plt.title(\"MinMax Scaler\")\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "5047339c-8c86-4513-91b3-684462cf3acb", "metadata": {}, "source": [ "#### 2.3.2.3 Using Box-Cox Transformation\n", "- A [Box Cox](https://rss.onlinelibrary.wiley.com/doi/10.1111/j.2517-6161.1964.tb00553.x) transformation is a transformation of a non-normal dependent variables into a normal shape. \n", "- Normality is an important assumption for many statistical techniques; if your data isn’t normal, applying a Box-Cox means that you are able to run a broader number of tests.\n", "- The Box Cox transformation is named after statisticians George Box and Sir David Roxbee Cox who collaborated on a 1964 paper and developed the technique.\n", "- BoxCox can only be applied to stricly positive values" ] }, { "cell_type": "code", "execution_count": 12, "id": "cc8af290-49fe-4930-9cee-d9429326a709", "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", "
OzoneSolar.RWindTempMonthDay
00.1907920.029277-0.696409-1.155384-1.441568-1.940899
10.002915-0.793127-0.513633-0.682490-1.441568-1.732923
2-1.313959-0.4420260.772939-0.481905-1.441568-1.553227
3-0.8795301.4776770.480592-1.587927-1.441568-1.389739
40.231016-0.0177991.209918-2.054572-1.441568-1.237337
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 0.190792 0.029277 -0.696409 -1.155384 -1.441568 -1.940899\n", "1 0.002915 -0.793127 -0.513633 -0.682490 -1.441568 -1.732923\n", "2 -1.313959 -0.442026 0.772939 -0.481905 -1.441568 -1.553227\n", "3 -0.879530 1.477677 0.480592 -1.587927 -1.441568 -1.389739\n", "4 0.231016 -0.017799 1.209918 -2.054572 -1.441568 -1.237337" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.preprocessing import power_transform\n", "data_BxCx = pd.DataFrame(power_transform(data3,method=\"box-cox\"), columns = data3.columns)\n", "data_BxCx.head()" ] }, { "cell_type": "code", "execution_count": 13, "id": "ee9ec7ba-c10a-456d-a19b-999c974d27d0", "metadata": {}, "outputs": [ { "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.hist(data_df['Temp'], bins=20)\n", "plt.title(\"Original\")\n", "plt.show()\n", "\n", "plt.hist(data_BxCx['Temp'], bins=20)\n", "plt.title(\"Box-Cox Scaler\")\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "c24cfdc9-cd04-48dd-a514-e7f75d3fe5a4", "metadata": {}, "source": [ "#### 2.3.2.4 Using Yeo Johnson Transformation\n", "While BoxCox only works with positive value, a more recent transformation method [Yeo Johnson](https://www.jstor.org/stable/2673623) can transform both positive and negative values." ] }, { "cell_type": "code", "execution_count": 14, "id": "b615ec00-4107-45b7-b850-eb9ee70b3b8a", "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", "
OzoneSolar.RWindTempMonthDay
00.1967150.027324-0.697609-1.155302-1.437509-1.891431
10.007778-0.795065-0.514183-0.682500-1.437509-1.709416
2-1.326287-0.4443160.774562-0.481950-1.437509-1.543376
3-0.8854311.4808860.482270-1.587780-1.437509-1.388286
40.237096-0.0198231.210699-2.054440-1.437509-1.241399
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 0.196715 0.027324 -0.697609 -1.155302 -1.437509 -1.891431\n", "1 0.007778 -0.795065 -0.514183 -0.682500 -1.437509 -1.709416\n", "2 -1.326287 -0.444316 0.774562 -0.481950 -1.437509 -1.543376\n", "3 -0.885431 1.480886 0.482270 -1.587780 -1.437509 -1.388286\n", "4 0.237096 -0.019823 1.210699 -2.054440 -1.437509 -1.241399" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data_yeo_johnson = pd.DataFrame(power_transform(data3,method=\"yeo-johnson\"))\n", "data_yeo_johnson.columns = data3.columns\n", "data_yeo_johnson.head()" ] }, { "cell_type": "code", "execution_count": 15, "id": "7df79828-c09b-4934-aab4-57fb5d33b593", "metadata": {}, "outputs": [ { "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.hist(data_df['Temp'], bins=20)\n", "plt.title(\"Original\")\n", "plt.show()\n", "\n", "plt.hist(data_yeo_johnson['Temp'], bins=20)\n", "plt.title(\"Yeo Johnson Scaler\")\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "07f0b47e-aee5-4d30-89c3-fac55b0e5059", "metadata": { "tags": [] }, "source": [ "## Feature engineering / selection" ] }, { "cell_type": "markdown", "id": "c1aa5c5c-0cc7-4a5a-b2b4-9f7743c0c037", "metadata": { "tags": [] }, "source": [ "Transform the data to facilitate learning. Many techniques, and we will only scratch the surface: \n", "* Manual feature engineering\n", "* SelectKBest\n", "* PCA\n", "* One-Hot Encoder" ] }, { "cell_type": "code", "execution_count": 16, "id": "e1664832-c8ee-4d64-a0ff-334f78ea4b9c", "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", "
OzoneSolar.RWindTempMonthDay
041.00000190.0000007.467.05.01.0
136.00000118.0000008.072.05.02.0
212.00000149.00000012.674.05.03.0
318.00000313.00000011.562.05.04.0
442.12931185.93150714.356.05.05.0
\n", "
" ], "text/plain": [ " Ozone Solar.R Wind Temp Month Day\n", "0 41.00000 190.000000 7.4 67.0 5.0 1.0\n", "1 36.00000 118.000000 8.0 72.0 5.0 2.0\n", "2 12.00000 149.000000 12.6 74.0 5.0 3.0\n", "3 18.00000 313.000000 11.5 62.0 5.0 4.0\n", "4 42.12931 185.931507 14.3 56.0 5.0 5.0" ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data3.head()" ] }, { "cell_type": "code", "execution_count": 17, "id": "49fc2443-85c3-4fd5-9bef-597ae624f668", "metadata": { "tags": [] }, "outputs": [ { "data": { "text/plain": [ "((153, 5), (153,))" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "X, y = data3.iloc[:,1:], data3.iloc[:,0]\n", "X.shape, y.shape" ] }, { "cell_type": "markdown", "id": "1442cbb9-33ae-4832-afe0-16582c635a1b", "metadata": { "jp-MarkdownHeadingCollapsed": true, "tags": [] }, "source": [ "### Manual Feature Engineering" ] }, { "cell_type": "code", "execution_count": 18, "id": "13612db1-dff8-4594-9a16-38996100c6ae", "metadata": { "tags": [] }, "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", "
Solar.RWindTempMonthDaySolar.R / Temp
0190.0000007.467.05.01.02.835821
1118.0000008.072.05.02.01.638889
2149.00000012.674.05.03.02.013514
3313.00000011.562.05.04.05.048387
4185.93150714.356.05.05.03.320205
.....................
148193.0000006.970.09.026.02.757143
149145.00000013.277.09.027.01.883117
150191.00000014.375.09.028.02.546667
151131.0000008.076.09.029.01.723684
152223.00000011.568.09.030.03.279412
\n", "

153 rows × 6 columns

\n", "
" ], "text/plain": [ " Solar.R Wind Temp Month Day Solar.R / Temp\n", "0 190.000000 7.4 67.0 5.0 1.0 2.835821\n", "1 118.000000 8.0 72.0 5.0 2.0 1.638889\n", "2 149.000000 12.6 74.0 5.0 3.0 2.013514\n", "3 313.000000 11.5 62.0 5.0 4.0 5.048387\n", "4 185.931507 14.3 56.0 5.0 5.0 3.320205\n", ".. ... ... ... ... ... ...\n", "148 193.000000 6.9 70.0 9.0 26.0 2.757143\n", "149 145.000000 13.2 77.0 9.0 27.0 1.883117\n", "150 191.000000 14.3 75.0 9.0 28.0 2.546667\n", "151 131.000000 8.0 76.0 9.0 29.0 1.723684\n", "152 223.000000 11.5 68.0 9.0 30.0 3.279412\n", "\n", "[153 rows x 6 columns]" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Let's add a NEW feature - a ratio of two of the measurements\n", "X['Solar.R / Temp'] = X['Solar.R'] / X['Temp']\n", "X" ] }, { "cell_type": "markdown", "id": "43b17605-bbf5-45ff-b4f8-9d1d9e2b0355", "metadata": { "jp-MarkdownHeadingCollapsed": true, "tags": [] }, "source": [ "### [Feature selection](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.feature_selection) " ] }, { "cell_type": "code", "execution_count": 19, "id": "3b732174-d39a-42c9-84f0-e9dc5b1a4c04", "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Top k features: ['Wind', 'Temp']\n" ] } ], "source": [ "# SelectKBest for selecting top-scoring features\n", "from sklearn.feature_selection import SelectKBest, f_regression\n", "\n", "# select the best 3 features for regression\n", "dim_red = SelectKBest(f_regression, k = 2)\n", "dim_red.fit(X, y)\n", "X_t = dim_red.transform(X)\n", "\n", "# Get back the selected columns\n", "selected = dim_red.get_support() # boolean values\n", "selected_names = X.columns[selected]\n", "\n", "print('Top k features: ', list(selected_names))" ] }, { "cell_type": "code", "execution_count": 20, "id": "066b99f0-7c92-4435-90f0-8f86c176e8e2", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Scores: [1.52612031e+01 5.92750302e+01 8.88983834e+01 3.43229390e+00\n", " 1.94731074e-02 3.35729306e+00]\n", "New shape: (153, 2)\n" ] } ], "source": [ "# Show scores, features selected and new shape\n", "print('Scores:', dim_red.scores_)\n", "print('New shape:', X_t.shape)" ] }, { "cell_type": "markdown", "id": "c6c2bce3-1cbe-4f17-b596-9e914ae3ad00", "metadata": {}, "source": [ "**Note on scoring function selection in `SelectKBest` tranformations:**\n", "* For regression - [f_regression](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.f_regression.html#sklearn.feature_selection.f_regression)\n", "* For classification - [chi2](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.chi2.html#sklearn.feature_selection.chi2), [f_classif](http://scikit-learn.org/stable/modules/generated/sklearn.feature_selection.f_classif.html#sklearn.feature_selection.f_classif)\n" ] }, { "cell_type": "markdown", "id": "0f2c0803-d5e3-4c5d-b1b4-04c71300e4c6", "metadata": { "jp-MarkdownHeadingCollapsed": true, "tags": [] }, "source": [ "### Principal component analysis (aka PCA)\n", "* Reduces dimensions (number of features), based on what information explains the most variance (or signal)\n", "* Considered unsupervised learning\n", "* Useful for very large feature space (e.g. say the botanist in charge of the iris dataset measured 100 more parts of the flower and thus there were 104 columns instead of 4)\n", "* More about PCA on wikipedia [here](https://en.wikipedia.org/wiki/Principal_component_analysis)" ] }, { "cell_type": "code", "execution_count": 21, "id": "510ce77f-d34b-4bd6-b83a-6e8107d351ea", "metadata": { "tags": [] }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# PCA for dimensionality reduction\n", "\n", "from sklearn import decomposition\n", "from sklearn import datasets\n", "\n", "digits = datasets.load_digits()\n", "\n", "X, y = digits.data, digits.target\n", "\n", "# perform principal component analysis\n", "pca = decomposition.PCA(.95)\n", "pca.fit(X)\n", "X_t = pca.transform(X)\n", "(X_t[:, 0])\n", "\n", "# import numpy and matplotlib for plotting (and set some stuff)\n", "import numpy as np\n", "np.set_printoptions(suppress=True)\n", "import matplotlib.pyplot as plt\n", "%matplotlib inline\n", "\n", "# let's separate out data based on first two principle components\n", "x1, x2 = X_t[:, 0], X_t[:, 1]\n", "\n", "\n", "# please don't worry about details of the plotting below \n", "c1 = np.array(list('rbg')) # colors\n", "classes = digits.target_names[y] \n", "for (i, cla) in enumerate(set(classes)):\n", " xc = [p for (j, p) in enumerate(x1) if classes[j] == cla]\n", " yc = [p for (j, p) in enumerate(x2) if classes[j] == cla]\n", " plt.scatter(xc, yc, label = cla)\n", " plt.ylabel('Principal Component 2')\n", " plt.xlabel('Principal Component 1')\n", "plt.legend(loc = 4)" ] }, { "cell_type": "markdown", "id": "8f86dd33-fe78-4137-899d-3459ca17950b", "metadata": {}, "source": [ "See scikit-learn's excellent tutorial on feature selection [here](http://scikit-learn.org/stable/modules/feature_selection.html)" ] }, { "cell_type": "markdown", "id": "d01843e2-bdb3-41d4-aa25-36d0e4531103", "metadata": {}, "source": [ "### One Hot Encoding\n", "* It's an operation on feature labels - a method of dummying variable\n", "* Expands the feature space by nature of transform - later this can be processed further with a dimensionality reduction (the dummied variables are now their own features)\n", "* FYI: One hot encoding variables is needed for python ML module `tenorflow`\n", "* Can do this with `pandas` method or a `sklearn` one-hot-encoder system" ] }, { "cell_type": "markdown", "id": "1d932024-a1f2-4b7c-8f5c-5fc27b8578a2", "metadata": {}, "source": [ "#### `pandas` method" ] }, { "cell_type": "code", "execution_count": 22, "id": "f535768c-a9fa-4c30-836a-6353de72499e", "metadata": { "tags": [] }, "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", "
sepal length (cm)sepal width (cm)petal length (cm)petal width (cm)target_name
05.13.51.40.2setosa
14.93.01.40.2setosa
24.73.21.30.2setosa
34.63.11.50.2setosa
45.03.61.40.2setosa
..................
1456.73.05.22.3virginica
1466.32.55.01.9virginica
1476.53.05.22.0virginica
1486.23.45.42.3virginica
1495.93.05.11.8virginica
\n", "

150 rows × 5 columns

\n", "
" ], "text/plain": [ " sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) \\\n", "0 5.1 3.5 1.4 0.2 \n", "1 4.9 3.0 1.4 0.2 \n", "2 4.7 3.2 1.3 0.2 \n", "3 4.6 3.1 1.5 0.2 \n", "4 5.0 3.6 1.4 0.2 \n", ".. ... ... ... ... \n", "145 6.7 3.0 5.2 2.3 \n", "146 6.3 2.5 5.0 1.9 \n", "147 6.5 3.0 5.2 2.0 \n", "148 6.2 3.4 5.4 2.3 \n", "149 5.9 3.0 5.1 1.8 \n", "\n", " target_name \n", "0 setosa \n", "1 setosa \n", "2 setosa \n", "3 setosa \n", "4 setosa \n", ".. ... \n", "145 virginica \n", "146 virginica \n", "147 virginica \n", "148 virginica \n", "149 virginica \n", "\n", "[150 rows x 5 columns]" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Dummy variables with pandas built-in function\n", "\n", "import pandas as pd\n", "from sklearn import datasets\n", "\n", "iris = datasets.load_iris()\n", "X, y = iris.data, iris.target\n", "\n", "# Convert to dataframe and add a column with actual iris species name\n", "data = pd.DataFrame(X, columns = iris.feature_names)\n", "data['target_name'] = iris.target_names[y]\n", "data" ] }, { "cell_type": "code", "execution_count": 23, "id": "e794b056-53e5-46ba-9c78-dde2db7ed5c2", "metadata": { "tags": [] }, "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", "
sepal length (cm)sepal width (cm)petal length (cm)petal width (cm)target_name_setosatarget_name_versicolortarget_name_virginica
1215.62.84.92.0001
756.63.04.41.4010
845.43.04.51.5010
195.13.81.50.3100
925.82.64.01.2010
185.73.81.70.3100
1064.92.54.51.7001
344.93.11.50.2100
1006.33.36.02.5001
507.03.24.71.4010
\n", "
" ], "text/plain": [ " sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) \\\n", "121 5.6 2.8 4.9 2.0 \n", "75 6.6 3.0 4.4 1.4 \n", "84 5.4 3.0 4.5 1.5 \n", "19 5.1 3.8 1.5 0.3 \n", "92 5.8 2.6 4.0 1.2 \n", "18 5.7 3.8 1.7 0.3 \n", "106 4.9 2.5 4.5 1.7 \n", "34 4.9 3.1 1.5 0.2 \n", "100 6.3 3.3 6.0 2.5 \n", "50 7.0 3.2 4.7 1.4 \n", "\n", " target_name_setosa target_name_versicolor target_name_virginica \n", "121 0 0 1 \n", "75 0 1 0 \n", "84 0 1 0 \n", "19 1 0 0 \n", "92 0 1 0 \n", "18 1 0 0 \n", "106 0 0 1 \n", "34 1 0 0 \n", "100 0 0 1 \n", "50 0 1 0 " ] }, "execution_count": 23, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.get_dummies(data, prefix = ['target_name'])\n", "df.sample(10)" ] }, { "cell_type": "markdown", "id": "85f52403-2510-4b7c-8e0a-e64560f62500", "metadata": {}, "source": [ "#### `sklearn` method" ] }, { "cell_type": "code", "execution_count": 24, "id": "ed965cec-762d-4a91-8ab7-5060efb1156f", "metadata": { "tags": [] }, "outputs": [ { "data": { "text/plain": [ "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n", " 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n", " 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,\n", " 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,\n", " 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# OneHotEncoder for dummying variables\n", "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n", "from sklearn import datasets\n", "\n", "iris = datasets.load_iris()\n", "X, y = iris.data, iris.target\n", "\n", "# We encode both our categorical variable and it's labels\n", "enc = OneHotEncoder()\n", "label_enc = LabelEncoder() # remember the labels here\n", "\n", "# Encode labels (can use for discrete numerical values as well)\n", "data_label_encoded = label_enc.fit_transform(y)\n", "data_label_encoded" ] }, { "cell_type": "code", "execution_count": 25, "id": "3179a111-3523-490f-88d7-0c14c27e51e4", "metadata": { "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(150, 3)\n" ] }, { "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", "
012
01.00.00.0
11.00.00.0
21.00.00.0
31.00.00.0
41.00.00.0
\n", "
" ], "text/plain": [ " 0 1 2\n", "0 1.0 0.0 0.0\n", "1 1.0 0.0 0.0\n", "2 1.0 0.0 0.0\n", "3 1.0 0.0 0.0\n", "4 1.0 0.0 0.0" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Encode and \"dummy\" variables\n", "data_feature_one_hot_encoded = enc.fit_transform(y.reshape(-1, 1))\n", "print(data_feature_one_hot_encoded.shape)\n", "\n", "num_dummies = data_feature_one_hot_encoded.shape[1]\n", "df = pd.DataFrame(data_feature_one_hot_encoded.toarray(), columns = label_enc.inverse_transform(range(num_dummies)))\n", "\n", "df.head()" ] }, { "cell_type": "code", "execution_count": null, "id": "9188dcd4-52b5-4210-a41c-0ee1e816e9ba", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "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": "3.9.12" } }, "nbformat": 4, "nbformat_minor": 5 }