diff --git "a/your-code/main (6) - c\303\263pia.ipynb" "b/your-code/main (6) - c\303\263pia.ipynb" new file mode 100644 index 0000000..6083968 --- /dev/null +++ "b/your-code/main (6) - c\303\263pia.ipynb" @@ -0,0 +1,3076 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "id": "Wspp3b7n2aeZ" + }, + "source": [ + "# Before your start:\n", + "- Read the README.md file\n", + "- Comment as much as you can and use the resources in the README.md file\n", + "- Happy learning!" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "id": "0dFWp9pz2aed" + }, + "outputs": [], + "source": [ + "# Import your libraries:\n", + "\n", + "%matplotlib inline\n", + "\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "import pandas as pd\n", + "from scipy import stats\n", + "from sklearn.preprocessing import StandardScaler\n", + "from sklearn.cluster import KMeans\n", + "from sklearn.cluster import DBSCAN" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "BkXSG3_82aee" + }, + "source": [ + "# Challenge 1 - Import and Describe the Dataset\n", + "\n", + "In this lab, we will use a dataset containing information about customer preferences. We will look at how much each customer spends in a year on each subcategory in the grocery store and try to find similarities using clustering.\n", + "\n", + "The origin of the dataset is [here](https://archive.ics.uci.edu/ml/datasets/wholesale+customers)." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "id": "0gALCkd32aee" + }, + "outputs": [], + "source": [ + "# loading the data:\n", + "customers = pd.read_csv('/content/Wholesale customers data.csv')" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "pptq-ffq2aee" + }, + "source": [ + "#### Explore the dataset with mathematical and visualization techniques. What do you find?\n", + "\n", + "Checklist:\n", + "\n", + "* What does each column mean?\n", + "* Any categorical data to convert?\n", + "* Any missing data to remove?\n", + "* Column collinearity - any high correlations?\n", + "* Descriptive statistics - any outliers to remove?\n", + "* Column-wise data distribution - is the distribution skewed?\n", + "* Etc.\n", + "\n", + "Additional info: Over a century ago, an Italian economist named Vilfredo Pareto discovered that roughly 20% of the customers account for 80% of the typical retail sales. This is called the [Pareto principle](https://en.wikipedia.org/wiki/Pareto_principle). Check if this dataset displays this characteristic." + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + }, + "id": "6ta3to4K2aef", + "outputId": "1e81be42-68b8-4d0e-faa8-350d51cbcfbd" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery Frozen Detergents_Paper Delicassen\n", + "0 2 3 12669 9656 7561 214 2674 1338\n", + "1 2 3 7057 9810 9568 1762 3293 1776\n", + "2 2 3 6353 8808 7684 2405 3516 7844\n", + "3 1 3 13265 1196 4221 6404 507 1788\n", + "4 2 3 22615 5410 7198 3915 1777 5185" + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassen
023126699656756121426741338
123705798109568176232931776
223635388087684240535167844
313132651196422164045071788
4232261554107198391517775185
\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 5 + } + ], + "source": [ + "# Your code here:\n", + "customers.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "Msw9vbn_2aef", + "outputId": "953fd020-b962-4a28-bdfe-7b6ce68b6469" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "Channel int64\n", + "Region int64\n", + "Fresh int64\n", + "Milk int64\n", + "Grocery int64\n", + "Frozen int64\n", + "Detergents_Paper int64\n", + "Delicassen int64\n", + "dtype: object" + ] + }, + "metadata": {}, + "execution_count": 6 + } + ], + "source": [ + "# Your observations here\n", + "customers.dtypes" + ] + }, + { + "cell_type": "code", + "source": [ + "customers.isnull().sum()" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "76hZ9FgG25Od", + "outputId": "eb135c87-4b47-47d7-a5e7-634b78509bc9" + }, + "execution_count": 7, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "Channel 0\n", + "Region 0\n", + "Fresh 0\n", + "Milk 0\n", + "Grocery 0\n", + "Frozen 0\n", + "Detergents_Paper 0\n", + "Delicassen 0\n", + "dtype: int64" + ] + }, + "metadata": {}, + "execution_count": 7 + } + ] + }, + { + "cell_type": "code", + "source": [ + "corr = customers.corr()\n", + "corr" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 300 + }, + "id": "NOY17RXJ2-Ic", + "outputId": "6f05799e-ee44-4da0-d1db-4461b1c8f68d" + }, + "execution_count": 8, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery Frozen \\\n", + "Channel 1.000000 0.062028 -0.169172 0.460720 0.608792 -0.202046 \n", + "Region 0.062028 1.000000 0.055287 0.032288 0.007696 -0.021044 \n", + "Fresh -0.169172 0.055287 1.000000 0.100510 -0.011854 0.345881 \n", + "Milk 0.460720 0.032288 0.100510 1.000000 0.728335 0.123994 \n", + "Grocery 0.608792 0.007696 -0.011854 0.728335 1.000000 -0.040193 \n", + "Frozen -0.202046 -0.021044 0.345881 0.123994 -0.040193 1.000000 \n", + "Detergents_Paper 0.636026 -0.001483 -0.101953 0.661816 0.924641 -0.131525 \n", + "Delicassen 0.056011 0.045212 0.244690 0.406368 0.205497 0.390947 \n", + "\n", + " Detergents_Paper Delicassen \n", + "Channel 0.636026 0.056011 \n", + "Region -0.001483 0.045212 \n", + "Fresh -0.101953 0.244690 \n", + "Milk 0.661816 0.406368 \n", + "Grocery 0.924641 0.205497 \n", + "Frozen -0.131525 0.390947 \n", + "Detergents_Paper 1.000000 0.069291 \n", + "Delicassen 0.069291 1.000000 " + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassen
Channel1.0000000.062028-0.1691720.4607200.608792-0.2020460.6360260.056011
Region0.0620281.0000000.0552870.0322880.007696-0.021044-0.0014830.045212
Fresh-0.1691720.0552871.0000000.100510-0.0118540.345881-0.1019530.244690
Milk0.4607200.0322880.1005101.0000000.7283350.1239940.6618160.406368
Grocery0.6087920.007696-0.0118540.7283351.000000-0.0401930.9246410.205497
Frozen-0.202046-0.0210440.3458810.123994-0.0401931.000000-0.1315250.390947
Detergents_Paper0.636026-0.001483-0.1019530.6618160.924641-0.1315251.0000000.069291
Delicassen0.0560110.0452120.2446900.4063680.2054970.3909470.0692911.000000
\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 8 + } + ] + }, + { + "cell_type": "code", + "source": [ + "customers.describe().round(2)" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 300 + }, + "id": "nej17TqG3ATI", + "outputId": "707f8f99-085d-4ecc-e095-6f6db6e369c6" + }, + "execution_count": 9, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery Frozen \\\n", + "count 440.00 440.00 440.00 440.00 440.00 440.00 \n", + "mean 1.32 2.54 12000.30 5796.27 7951.28 3071.93 \n", + "std 0.47 0.77 12647.33 7380.38 9503.16 4854.67 \n", + "min 1.00 1.00 3.00 55.00 3.00 25.00 \n", + "25% 1.00 2.00 3127.75 1533.00 2153.00 742.25 \n", + "50% 1.00 3.00 8504.00 3627.00 4755.50 1526.00 \n", + "75% 2.00 3.00 16933.75 7190.25 10655.75 3554.25 \n", + "max 2.00 3.00 112151.00 73498.00 92780.00 60869.00 \n", + "\n", + " Detergents_Paper Delicassen \n", + "count 440.00 440.00 \n", + "mean 2881.49 1524.87 \n", + "std 4767.85 2820.11 \n", + "min 3.00 3.00 \n", + "25% 256.75 408.25 \n", + "50% 816.50 965.50 \n", + "75% 3922.00 1820.25 \n", + "max 40827.00 47943.00 " + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassen
count440.00440.00440.00440.00440.00440.00440.00440.00
mean1.322.5412000.305796.277951.283071.932881.491524.87
std0.470.7712647.337380.389503.164854.674767.852820.11
min1.001.003.0055.003.0025.003.003.00
25%1.002.003127.751533.002153.00742.25256.75408.25
50%1.003.008504.003627.004755.501526.00816.50965.50
75%2.003.0016933.757190.2510655.753554.253922.001820.25
max2.003.00112151.0073498.0092780.0060869.0040827.0047943.00
\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 9 + } + ] + }, + { + "cell_type": "code", + "source": [ + "columns_1 = customers.select_dtypes(include=['float64', 'int64'])\n", + "\n", + "\n", + "columns_1.hist(figsize=(12, 8), bins=20, grid=False)\n", + "plt.suptitle(y=1.02, fontsize=16)\n", + "plt.tight_layout()\n", + "plt.show()" + ], + "metadata": { + "id": "PEz1YHQy3Crc" + }, + "execution_count": null, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "tmVbYFwZ2aef" + }, + "source": [ + "# Challenge 2 - Data Cleaning and Transformation\n", + "\n", + "If your conclusion from the previous challenge is the data need cleaning/transformation, do it in the cells below. However, if your conclusion is the data need not be cleaned or transformed, feel free to skip this challenge. But if you do choose the latter, please provide rationale." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "L4DyRjIp2aef" + }, + "outputs": [], + "source": [ + "# Your code here" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "iRdVavYs2aeg" + }, + "outputs": [], + "source": [ + "# Your comment here" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "7Dm7S3042aeg" + }, + "source": [ + "# Challenge 3 - Data Preprocessing\n", + "\n", + "One problem with the dataset is the value ranges are remarkably different across various categories (e.g. `Fresh` and `Grocery` compared to `Detergents_Paper` and `Delicassen`). If you made this observation in the first challenge, you've done a great job! This means you not only completed the bonus questions in the previous Supervised Learning lab but also researched deep into [*feature scaling*](https://en.wikipedia.org/wiki/Feature_scaling). Keep on the good work!\n", + "\n", + "Diverse value ranges in different features could cause issues in our clustering. The way to reduce the problem is through feature scaling. We'll use this technique again with this dataset.\n", + "\n", + "#### We will use the `StandardScaler` from `sklearn.preprocessing` and scale our data. Read more about `StandardScaler` [here](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html#sklearn.preprocessing.StandardScaler).\n", + "\n", + "*After scaling your data, assign the transformed data to a new variable `customers_scale`.*" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 423 + }, + "id": "lYoH9ze52aeg", + "outputId": "ea065618-5af6-4458-8ac6-8ad8acd21170" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery Frozen \\\n", + "0 1.448652 0.590668 0.052933 0.523568 -0.041115 -0.589367 \n", + "1 1.448652 0.590668 -0.391302 0.544458 0.170318 -0.270136 \n", + "2 1.448652 0.590668 -0.447029 0.408538 -0.028157 -0.137536 \n", + "3 -0.690297 0.590668 0.100111 -0.624020 -0.392977 0.687144 \n", + "4 1.448652 0.590668 0.840239 -0.052396 -0.079356 0.173859 \n", + ".. ... ... ... ... ... ... \n", + "435 -0.690297 0.590668 1.401312 0.848446 0.850760 2.075222 \n", + "436 -0.690297 0.590668 2.155293 -0.592142 -0.757165 0.296561 \n", + "437 1.448652 0.590668 0.200326 1.314671 2.348386 -0.543380 \n", + "438 -0.690297 0.590668 -0.135384 -0.517536 -0.602514 -0.419441 \n", + "439 -0.690297 0.590668 -0.729307 -0.555924 -0.573227 -0.620094 \n", + "\n", + " Detergents_Paper Delicassen \n", + "0 -0.043569 -0.066339 \n", + "1 0.086407 0.089151 \n", + "2 0.133232 2.243293 \n", + "3 -0.498588 0.093411 \n", + "4 -0.231918 1.299347 \n", + ".. ... ... \n", + "435 -0.566831 0.241091 \n", + "436 -0.585519 0.291501 \n", + "437 2.511218 0.121456 \n", + "438 -0.569770 0.213046 \n", + "439 -0.504888 -0.522869 \n", + "\n", + "[440 rows x 8 columns]" + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassen
01.4486520.5906680.0529330.523568-0.041115-0.589367-0.043569-0.066339
11.4486520.590668-0.3913020.5444580.170318-0.2701360.0864070.089151
21.4486520.590668-0.4470290.408538-0.028157-0.1375360.1332322.243293
3-0.6902970.5906680.100111-0.624020-0.3929770.687144-0.4985880.093411
41.4486520.5906680.840239-0.052396-0.0793560.173859-0.2319181.299347
...........................
435-0.6902970.5906681.4013120.8484460.8507602.075222-0.5668310.241091
436-0.6902970.5906682.155293-0.592142-0.7571650.296561-0.5855190.291501
4371.4486520.5906680.2003261.3146712.348386-0.5433802.5112180.121456
438-0.6902970.590668-0.135384-0.517536-0.602514-0.419441-0.5697700.213046
439-0.6902970.590668-0.729307-0.555924-0.573227-0.620094-0.504888-0.522869
\n", + "

440 rows × 8 columns

\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 11 + } + ], + "source": [ + "# Your import here:\n", + "\n", + "from sklearn.preprocessing import StandardScaler\n", + "\n", + "# Your code here:\n", + "\n", + "\n", + "\n", + "columns_2 = customers.select_dtypes(include=['float64', 'int64'])\n", + "scaler = StandardScaler()\n", + "customers_scale = scaler.fit_transform(columns_2)\n", + "\n", + "customers_scale_df = pd.DataFrame(customers_scale, columns=columns_2.columns)\n", + "\n", + "customers_scale_df" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "N_uXwVD02aeh" + }, + "source": [ + "# Challenge 4 - Data Clustering with K-Means\n", + "\n", + "Now let's cluster the data with K-Means first. Initiate the K-Means model, then fit your scaled data. In the data returned from the `.fit` method, there is an attribute called `labels_` which is the cluster number assigned to each data record. What you can do is to assign these labels back to `customers` in a new column called `customers['labels']`. Then you'll see the cluster results of the original data." + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": { + "scrolled": true, + "colab": { + "base_uri": "https://localhost:8080/", + "height": 815 + }, + "id": "jijfKYQP2aeh", + "outputId": "34fdc3eb-ee71-44b0-df83-6ab36e5b0b61" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n", + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n" + ] + }, + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "" + ] + }, + "metadata": {}, + "execution_count": 12 + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ], + "source": [ + "# Your code here:\n", + "k_value = range (1, 11)\n", + "inercias=[]\n", + "\n", + "\n", + "for k in k_value:\n", + " k_means = KMeans(n_clusters=k, random_state=42)\n", + " k_means.fit(customers_scale_df)\n", + " inercias.append(k_means.inertia_)\n", + "\n", + "\n", + "plt.plot(k_value, inercias, marker = 'o')\n", + "plt.show" + ] + }, + { + "cell_type": "code", + "source": [ + "scaler = StandardScaler()\n", + "customers_scale = scaler.fit_transform(columns_2)\n", + "customers_scale_df = pd.DataFrame(customers_scale, columns=columns_2.columns)\n", + "k = 4\n", + "\n", + "k_means = KMeans(n_clusters=k, random_state=42)\n", + "\n", + "labels_clusters = k_means.fit_predict(customers_scale_df)\n", + "customers['labels'] = labels_clusters\n", + "\n", + "print(customers.head())" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "GI1cjWaF3iS4", + "outputId": "8d0d6780-007d-4a65-9628-d5c2a8252713" + }, + "execution_count": 13, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/sklearn/cluster/_kmeans.py:870: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n", + " warnings.warn(\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + " Channel Region Fresh Milk Grocery Frozen Detergents_Paper \\\n", + "0 2 3 12669 9656 7561 214 2674 \n", + "1 2 3 7057 9810 9568 1762 3293 \n", + "2 2 3 6353 8808 7684 2405 3516 \n", + "3 1 3 13265 1196 4221 6404 507 \n", + "4 2 3 22615 5410 7198 3915 1777 \n", + "\n", + " Delicassen labels \n", + "0 1338 2 \n", + "1 1776 2 \n", + "2 7844 2 \n", + "3 1788 0 \n", + "4 5185 2 \n" + ] + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "nqDM178h2aeh" + }, + "source": [ + "Count the values in `labels`." + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "aSAj9f1P2aeh", + "outputId": "ae64e510-675d-41ba-b060-9667e003fd1b" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "0 210\n", + "2 125\n", + "3 92\n", + "1 13\n", + "Name: labels, dtype: int64" + ] + }, + "metadata": {}, + "execution_count": 14 + } + ], + "source": [ + "# Your code here:\n", + "count_labels = customers['labels'].value_counts()\n", + "\n", + "count_labels" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "0ZXUi6Fe2aeh" + }, + "source": [ + "# Challenge 5 - Data Clustering with DBSCAN\n", + "\n", + "Now let's cluster the data using DBSCAN. Use `DBSCAN(eps=0.5)` to initiate the model, then fit your scaled data. In the data returned from the `.fit` method, assign the `labels_` back to `customers['labels_DBSCAN']`. Now your original data have two labels, one from K-Means and the other from DBSCAN." + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": { + "scrolled": true, + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "fHBp2pLe2aei", + "outputId": "21a9397f-fa93-4bad-c664-f6367a95dfcb" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + " Channel Region Fresh Milk Grocery Frozen Detergents_Paper \\\n", + "0 2 3 12669 9656 7561 214 2674 \n", + "1 2 3 7057 9810 9568 1762 3293 \n", + "2 2 3 6353 8808 7684 2405 3516 \n", + "3 1 3 13265 1196 4221 6404 507 \n", + "4 2 3 22615 5410 7198 3915 1777 \n", + "\n", + " Delicassen labels label_dbscan \n", + "0 1338 2 -1 \n", + "1 1776 2 -1 \n", + "2 7844 2 -1 \n", + "3 1788 0 1 \n", + "4 5185 2 -1 \n" + ] + } + ], + "source": [ + "# Your code here\n", + "eps_value = 0.5\n", + "\n", + "model_dbscan = DBSCAN(eps=eps_value)\n", + "\n", + "label_dbscan = model_dbscan.fit_predict(customers_scale_df)\n", + "\n", + "customers['label_dbscan'] = label_dbscan\n", + "\n", + "print(customers.head())" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "U8TZ1EWo2aei" + }, + "source": [ + "Count the values in `labels_DBSCAN`." + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "U4rtUe2_2aei", + "outputId": "ce1dc437-c57e-46e4-8bb2-287de42ed92f" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "-1 255\n", + " 1 130\n", + " 5 22\n", + " 4 7\n", + " 6 6\n", + " 0 5\n", + " 3 5\n", + " 2 5\n", + " 7 5\n", + "Name: label_dbscan, dtype: int64" + ] + }, + "metadata": {}, + "execution_count": 16 + } + ], + "source": [ + "# Your code here\n", + "count_labels_dbscan = customers['label_dbscan'].value_counts()\n", + "count_labels_dbscan" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "CCem06av2aei" + }, + "source": [ + "# Challenge 6 - Compare K-Means with DBSCAN\n", + "\n", + "Now we want to visually compare how K-Means and DBSCAN have clustered our data. We will create scatter plots for several columns. For each of the following column pairs, plot a scatter plot using `labels` and another using `labels_DBSCAN`. Put them side by side to compare. Which clustering algorithm makes better sense?\n", + "\n", + "Columns to visualize:\n", + "\n", + "* `Detergents_Paper` as X and `Milk` as y\n", + "* `Grocery` as X and `Fresh` as y\n", + "* `Frozen` as X and `Delicassen` as y" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "E1Ckzirj2aei" + }, + "source": [ + "Visualize `Detergents_Paper` as X and `Milk` as y by `labels` and `labels_DBSCAN` respectively" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 502 + }, + "id": "opcbYdbQ2aej", + "outputId": "f1d0c2c2-cf59-42f8-f379-a034dc0472ff" + }, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ], + "source": [ + "# Your code here:\n", + "fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(12, 5))\n", + "\n", + "axes[0].scatter(customers['Detergents_Paper'], customers['Milk'], c=customers['labels'], cmap='viridis')\n", + "\n", + "axes[1].scatter(customers['Detergents_Paper'], customers['Milk'], c=customers['label_dbscan'], cmap='viridis')\n", + "axes[1].set_title('DBSCAN Clustering')\n", + "axes[1].set_xlabel('Detergents_Paper')\n", + "axes[1].set_ylabel('Milk')\n", + "\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "qF3LAlvr2aej" + }, + "source": [ + "Visualize `Grocery` as X and `Fresh` as y by `labels` and `labels_DBSCAN` respectively" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 465 + }, + "id": "AM_pTtlO2aej", + "outputId": "fb81b797-dfe2-4c93-adf2-114cd95bb95b" + }, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ], + "source": [ + "# Your code here:\n", + "fig, axs = plt.subplots(1, 2, figsize=(12, 5))\n", + "\n", + "axs[0].scatter(customers['Grocery'], customers['Fresh'], c=customers['labels'], cmap='viridis', label='labels')\n", + "\n", + "axs[0].set_xlabel('Grocery')\n", + "axs[0].set_ylabel('Fresh')\n", + "\n", + "axs[1].scatter(customers['Grocery'], customers['Fresh'], c=customers['label_dbscan'], cmap='viridis', label='label_dbscan')\n", + "\n", + "axs[1].set_xlabel('Grocery')\n", + "axs[1].set_ylabel('Fresh')\n", + "\n", + "\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Il2IoXUj2aej" + }, + "source": [ + "Visualize `Frozen` as X and `Delicassen` as y by `labels` and `labels_DBSCAN` respectively" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 468 + }, + "id": "caPpVWXj2aek", + "outputId": "4ad01650-9b23-4951-988e-43c47b7bae8c" + }, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/8AAAHDCAYAAAByJaqpAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACN1UlEQVR4nOzdd3gU5doG8HtmtqRuKiShd5AuIBARUQkGQRRBxXIUFVQUVMQDiudYsOGxN6x8il3EgiIIYmgCQWroRSAQWhIgyW7qlpn3+2PJwpLsprDJLsv9u649sjvvzDyzcnz2mXmLJIQQICIiIiIiIqKgJfs7ACIiIiIiIiKqWyz+iYiIiIiIiIIci38iIiIiIiKiIMfin4iIiIiIiCjIsfgnIiIiIiIiCnIs/omIiIiIiIiCHIt/IiIiIiIioiDH4p+IiIiIiIgoyLH4JyIiIiIiIgpyLP6JiIiIiIiIgpxfi/9nn30WkiS5vTp06ODaXlZWhvHjxyMuLg4REREYOXIkcnJy3I6RlZWFoUOHIiwsDA0bNsTkyZPhcDjc2ixbtgw9evSA0WhEmzZtMGvWrAqxzJgxAy1atEBISAj69OmDtWvX1sk1ExERXUiY64mIiAKDzt8BdOrUCX/++afrvU53OqRHH30U8+fPx5w5cxAVFYUJEyZgxIgRWLVqFQBAVVUMHToUiYmJWL16NY4dO4Y777wTer0eL730EgAgMzMTQ4cOxbhx4/D1118jLS0NY8eORVJSElJTUwEAs2fPxqRJk/Dhhx+iT58+eOutt5Camordu3ejYcOG1boOTdNw9OhRREZGQpIkX309REREtSaEQGFhIRo1agRZ9t/9fuZ6IiKiulGjXC/86JlnnhHdunWrdFtBQYHQ6/Vizpw5rs927twpAIj09HQhhBALFiwQsiyL7OxsV5sPPvhAmEwmYbVahRBCTJkyRXTq1Mnt2KNGjRKpqamu97179xbjx493vVdVVTRq1EhMnz692tdy6NAhAYAvvvjiiy++Au516NChauczX2Ou54svvvjii6+6f1Un1/v9yf8///yDRo0aISQkBMnJyZg+fTqaNWuGDRs2wG63IyUlxdW2Q4cOaNasGdLT09G3b1+kp6ejS5cuSEhIcLVJTU3FAw88gO3bt+Piiy9Genq62zHK20ycOBEAYLPZsGHDBkydOtW1XZZlpKSkID093WPcVqsVVqvV9V4IAQA4dOgQTCbTOX0nREREvmCxWNC0aVNERkb6NQ7meiIiorpRk1zv1+K/T58+mDVrFtq3b49jx45h2rRp6N+/P7Zt24bs7GwYDAZER0e77ZOQkIDs7GwAQHZ2ttuPgfLt5du8tbFYLCgtLUV+fj5UVa20za5duzzGPn36dEybNq3C5yaTiT8IiIgooPizizpzPRERUd2rTq73a/F/zTXXuP7ctWtX9OnTB82bN8f333+P0NBQP0ZWtalTp2LSpEmu9+V3XIiIiOg05noiIqLAEFBL/UVHR6Ndu3bYu3cvEhMTYbPZUFBQ4NYmJycHiYmJAIDExMQKMwKXv6+qjclkQmhoKOLj46EoSqVtyo9RGaPR6LrzzycARERE1cNcT0RE5B8BVfwXFRVh3759SEpKQs+ePaHX65GWlubavnv3bmRlZSE5ORkAkJycjK1btyI3N9fVZvHixTCZTOjYsaOrzZnHKG9TfgyDwYCePXu6tdE0DWlpaa42RERE5BvM9URERH5S7Slu68Bjjz0mli1bJjIzM8WqVatESkqKiI+PF7m5uUIIIcaNGyeaNWsmlixZItavXy+Sk5NFcnKya3+HwyE6d+4srr76apGRkSEWLlwoGjRoIKZOnepqs3//fhEWFiYmT54sdu7cKWbMmCEURRELFy50tfnuu++E0WgUs2bNEjt27BD33XefiI6OdptZuCpms1kAEGaz2QffDBER0bkLhNzEXE9ERFR3apKb/Fr8jxo1SiQlJQmDwSAaN24sRo0aJfbu3evaXlpaKh588EERExMjwsLCxA033CCOHTvmdowDBw6Ia665RoSGhor4+Hjx2GOPCbvd7tZm6dKlonv37sJgMIhWrVqJzz77rEIs7777rmjWrJkwGAyid+/eYs2aNTW6Fv4gICKiQBMIuYm5noiIqO7UJDdJQpxat4bOicViQVRUFMxmM8cEEhFRQGBu8i1+n0REFGhqkpsCasw/EREREREREfkei38iIiIiIiKiIMfin4iIiIiIiCjI6fwdABFRoCu12fHH+j3YvO8oJAm4pH1TXNW9DQx6/ieUiIgoGGiahowl27Dy57WwllrRolMzXD16AKLiOb8HBQ/+ciUi8mLz/qOY+P4vMBeXQZElAMDPq7YhPioc7024Ae2aNPBzhERERHQu8nPN+O+107Fn/T4oOgWAgKYJfPrk15j0yQMYdOcAf4dI5BPs9k9E5EF2XiHGv/MTCkusAABVE1A15wIpeYUlGPfWDzAXl/kzRCIiIjoHQgg8Nexl7M3IBACoDhWqQ4PQBBx2Fa/c/R42Ldnq5yiJfIPFPxGRB3NWbEaZ3QGtkhVRNU3AXFKGX9O3+yEyIiIi8oWMpduwe91eaA6t0u2yLOPb6T/Xc1REdYPFPxGRB39s2ANNq1j4lxMC+HPDnnqMiIiIiHxp9dx1p7r6V05TNWxK24pS9vSjIMDin4jIgzKbvco2pTZHPURCREREdaHs1NC+qtjLqv5NQBToWPwTEXnQvkkD1yR/lVFkCe2bcsI/IiKi81XLLs2gqZV3+S8X3TAKETHh9RQRUd1h8U9E5MHNV3R3TfBXGVUTuPHyrvUYEREREfnSoDsHQG/0vACaJEu47oFUyDLLJjr/8W8xEZEH/Tu3xA39OgMAznz+L0vOd3enXoJurRr5ITIiIiLyhciYCPz70/GQJAmy4l4aybKEDr3b4KbJ1/kpOiLf8nybi4joAidJEv57ewq6tEzCl39uQGZ2HgCgbeN4jL66F1J7tfdzhERERHSurrylH+IaxeDb6T9jwx8ZEAKISYjCdQ8Oxo2PDUNImNHfIRL5BIt/IiIvJEnC8H6dcf2lnVBcZoMkSQgPMfg7LCIiIvKhrpd3RNfLO8JaaoXd6kCYKZRd/SnosPgnIqoGSZIQEco7/0RERMHMGGqEkfmeghRvZxEREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQC5ji/+WXX4YkSZg4caLrs7KyMowfPx5xcXGIiIjAyJEjkZOT47ZfVlYWhg4dirCwMDRs2BCTJ0+Gw+Fwa7Ns2TL06NEDRqMRbdq0waxZsyqcf8aMGWjRogVCQkLQp08frF27ti4uk4iI6ILGfE9EROQfAVH8r1u3Dh999BG6du3q9vmjjz6KefPmYc6cOVi+fDmOHj2KESNGuLarqoqhQ4fCZrNh9erV+PzzzzFr1iw8/fTTrjaZmZkYOnQorrzySmRkZGDixIkYO3YsFi1a5Goze/ZsTJo0Cc888ww2btyIbt26ITU1Fbm5uXV/8URERBcI5nsiIiI/En5WWFgo2rZtKxYvXiwGDBggHnnkESGEEAUFBUKv14s5c+a42u7cuVMAEOnp6UIIIRYsWCBkWRbZ2dmuNh988IEwmUzCarUKIYSYMmWK6NSpk9s5R40aJVJTU13ve/fuLcaPH+96r6qqaNSokZg+fXq1r8NsNgsAwmw2V//iiYiI6lAg5aZgyPeB9H0SEREJUbPc5Pcn/+PHj8fQoUORkpLi9vmGDRtgt9vdPu/QoQOaNWuG9PR0AEB6ejq6dOmChIQEV5vU1FRYLBZs377d1ebsY6emprqOYbPZsGHDBrc2siwjJSXF1aYyVqsVFovF7UVERESVOx/zPXM9EREFE50/T/7dd99h48aNWLduXYVt2dnZMBgMiI6Odvs8ISEB2dnZrjZn/hAo316+zVsbi8WC0tJS5OfnQ1XVStvs2rXLY+zTp0/HtGnTqnehREREF7DzNd8z1xMRUTDx25P/Q4cO4ZFHHsHXX3+NkJAQf4VRa1OnToXZbHa9Dh065O+QiIiIAs75nO+Z64mIKJj4rfjfsGEDcnNz0aNHD+h0Ouh0OixfvhzvvPMOdDodEhISYLPZUFBQ4LZfTk4OEhMTAQCJiYkVZgMuf19VG5PJhNDQUMTHx0NRlErblB+jMkajESaTye1FRERE7s7nfM9cT0REwcRvxf/AgQOxdetWZGRkuF69evXC7bff7vqzXq9HWlqaa5/du3cjKysLycnJAIDk5GRs3brVbZbexYsXw2QyoWPHjq42Zx6jvE35MQwGA3r27OnWRtM0pKWludoQERFR7TDfExERBQa/jfmPjIxE586d3T4LDw9HXFyc6/MxY8Zg0qRJiI2NhclkwkMPPYTk5GT07dsXAHD11VejY8eOuOOOO/DKK68gOzsb//3vfzF+/HgYjUYAwLhx4/Dee+9hypQpuOeee7BkyRJ8//33mD9/vuu8kyZNwujRo9GrVy/07t0bb731FoqLi3H33XfX07dBREQUnJjviYiIAoNfJ/yryptvvglZljFy5EhYrVakpqbi/fffd21XFAW//fYbHnjgASQnJyM8PByjR4/Gc88952rTsmVLzJ8/H48++ijefvttNGnSBDNnzkRqaqqrzahRo3D8+HE8/fTTyM7ORvfu3bFw4cIKkwIRERGR7zHfExER1T1JCCH8HUQwsFgsiIqKgtls5phAIiIKCMxNvsXvk4iIAk1NcpPfxvwTERERERERUf1g8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Fj8ExEREREREQU5Fv9EREREREREQY7FPxEREREREVGQY/FPREREREREFORY/BMREREREREFORb/REREREREREGOxT8RERERERFRkGPxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBjsU/ERERERERUZBj8U9EREREREQU5Pxa/H/wwQfo2rUrTCYTTCYTkpOT8fvvv7u2l5WVYfz48YiLi0NERARGjhyJnJwct2NkZWVh6NChCAsLQ8OGDTF58mQ4HA63NsuWLUOPHj1gNBrRpk0bzJo1q0IsM2bMQIsWLRASEoI+ffpg7dq1dXLNREREFxLmeiIiosDg1+K/SZMmePnll7FhwwasX78eV111Fa6//nps374dAPDoo49i3rx5mDNnDpYvX46jR49ixIgRrv1VVcXQoUNhs9mwevVqfP7555g1axaefvppV5vMzEwMHToUV155JTIyMjBx4kSMHTsWixYtcrWZPXs2Jk2ahGeeeQYbN25Et27dkJqaitzc3Pr7MoiIiIIQcz0REVGAEAEmJiZGzJw5UxQUFAi9Xi/mzJnj2rZz504BQKSnpwshhFiwYIGQZVlkZ2e72nzwwQfCZDIJq9UqhBBiypQpolOnTm7nGDVqlEhNTXW97927txg/frzrvaqqolGjRmL69OnVjttsNgsAwmw21+yCiYiI6kig5ibmeiIiIt+oSW4KmDH/qqriu+++Q3FxMZKTk7FhwwbY7XakpKS42nTo0AHNmjVDeno6ACA9PR1dunRBQkKCq01qaiosFovriUJ6errbMcrblB/DZrNhw4YNbm1kWUZKSoqrTWWsVissFovbi4iIiDxjriciIvIfvxf/W7duRUREBIxGI8aNG4eff/4ZHTt2RHZ2NgwGA6Kjo93aJyQkIDs7GwCQnZ3t9mOgfHv5Nm9tLBYLSktLceLECaiqWmmb8mNUZvr06YiKinK9mjZtWqvrJyIiCnbM9URERP7n9+K/ffv2yMjIwN9//40HHngAo0ePxo4dO/wdVpWmTp0Ks9nseh06dMjfIREREQUk5noiIiL/0/k7AIPBgDZt2gAAevbsiXXr1uHtt9/GqFGjYLPZUFBQ4PZEICcnB4mJiQCAxMTECjP1ls8QfGabs2cNzsnJgclkQmhoKBRFgaIolbYpP0ZljEYjjEZj7S6aiIjoAsJcT0RE5H9+f/J/Nk3TYLVa0bNnT+j1eqSlpbm27d69G1lZWUhOTgYAJCcnY+vWrW4z9S5evBgmkwkdO3Z0tTnzGOVtyo9hMBjQs2dPtzaapiEtLc3VhoiIiHyHuZ6IiMgP6mECQo+eeOIJsXz5cpGZmSm2bNkinnjiCSFJkvjjjz+EEEKMGzdONGvWTCxZskSsX79eJCcni+TkZNf+DodDdO7cWVx99dUiIyNDLFy4UDRo0EBMnTrV1Wb//v0iLCxMTJ48WezcuVPMmDFDKIoiFi5c6Grz3XffCaPRKGbNmiV27Ngh7rvvPhEdHe02s3BVOAMwEREFmkDITcz1REREdacmucmvxf8999wjmjdvLgwGg2jQoIEYOHCg68eAEEKUlpaKBx98UMTExIiwsDBxww03iGPHjrkd48CBA+Kaa64RoaGhIj4+Xjz22GPCbre7tVm6dKno3r27MBgMolWrVuKzzz6rEMu7774rmjVrJgwGg+jdu7dYs2ZNja6FPwiIiCjQBEJuYq4nIiKqOzXJTZIQQviv30HwsFgsiIqKgtlshslk8nc4REREzE0+xu+TiIgCTU1yU8CN+SciIiIiIiIi32LxT0RERERERBTkWPwTERERERERBTkW/0RERERERERBTlfbHdPS0pCWlobc3Fxomua27dNPPz3nwIiIiMj/mO+JiIiCQ62K/2nTpuG5555Dr169kJSUBEmSfB0XERER+RnzPRERUfCoVfH/4YcfYtasWbjjjjt8HQ8REREFCOZ7IiKi4FGrMf82mw2XXnqpr2MhIiKiAMJ8T0REFDxqVfyPHTsW33zzja9jISIiogDCfE9ERBQ8atXtv6ysDB9//DH+/PNPdO3aFXq93m37G2+84ZPgiIiIyH+Y74mIiIJHrYr/LVu2oHv37gCAbdu2uW3jZEBERETBgfmeiIgoeNSq+F+6dKmv4yAiIqIAw3xPREQUPGo15r/c3r17sWjRIpSWlgIAhBA+CYqIiIgCB/M9ERHR+a9Wxf/JkycxcOBAtGvXDkOGDMGxY8cAAGPGjMFjjz3m0wCJiIjIP5jviYiIgketiv9HH30Uer0eWVlZCAsLc30+atQoLFy40GfBERERkf8w3xMREQWPWo35/+OPP7Bo0SI0adLE7fO2bdvi4MGDPgmMiIiI/Iv5noiIKHjU6sl/cXGx2xOAcnl5eTAajeccFBEREfkf8z0REVHwqFXx379/f3zxxReu95IkQdM0vPLKK7jyyit9FhwRERH5D/M9ERFR8KhVt/9XXnkFAwcOxPr162Gz2TBlyhRs374deXl5WLVqla9jJCIiIj9gviciIgoetXry37lzZ+zZsweXXXYZrr/+ehQXF2PEiBHYtGkTWrdu7esYiYiIyA+Y74mIiIKHJLhYr09YLBZERUXBbDbDZDL5OxwiIiLmJh/j90lERIGmJrmpVk/+Fy5ciJUrV7rez5gxA927d8dtt92G/Pz82hySiIiIAgzzPRERUfCoVfE/efJkWCwWAMDWrVsxadIkDBkyBJmZmZg0aZJPAyQiIiL/YL4nIiIKHrWa8C8zMxMdO3YEAPz4448YNmwYXnrpJWzcuBFDhgzxaYBERETkH8z3REREwaNWT/4NBgNKSkoAAH/++SeuvvpqAEBsbKzrCQERERGd35jviYiIgketnvxfdtllmDRpEvr164e1a9di9uzZAIA9e/agSZMmPg2QiIiI/IP5noiIKHjU6sn/e++9B51Ohx9++AEffPABGjduDAD4/fffMXjwYJ8GSERERP7BfE9ERBQ8uNSfj3D5HyIiCjTMTb7F75OIiAJNnS/1t3HjRmzdutX1/pdffsHw4cPx5JNPwmaz1eaQREREFGCY74mIiIJHrYr/+++/H3v27AEA7N+/H7fccgvCwsIwZ84cTJkyxacBEhERkX8w3xMREQWPWhX/e/bsQffu3QEAc+bMweWXX45vvvkGs2bNwo8//ujL+IiIiMhPmO+JiIiCR62KfyEENE0D4Fz6p3yt36ZNm+LEiRO+i46IiIj8hvmeiIgoeNSq+O/VqxdeeOEFfPnll1i+fDmGDh0KAMjMzERCQoJPAyQiIiL/YL4nIiIKHrUq/t966y1s3LgREyZMwH/+8x+0adMGAPDDDz/g0ksv9WmARERE5B/M90RERMHDp0v9lZWVQVEU6PV6Xx3yvMHlf4iIKNDUVW66UPM9cz0REQWamuQmnS9PHBIS4svDERERUQBiviciIjr/1Kr4V1UVb775Jr7//ntkZWVVWOs3Ly/PJ8ERERGR/zDfExERBY9ajfmfNm0a3njjDYwaNQpmsxmTJk3CiBEjIMsynn32WR+HSERERP7AfE9ERBQ8alX8f/311/jkk0/w2GOPQafT4dZbb8XMmTPx9NNPY82aNb6OkYiIiPyA+Z6IiCh41Kr4z87ORpcuXQAAERERMJvNAIBrr70W8+fP9110RERE5DfM90RERMGjVsV/kyZNcOzYMQBA69at8ccffwAA1q1bB6PR6LvoiIiIyG+Y74mIiIJHrYr/G264AWlpaQCAhx56CE899RTatm2LO++8E/fcc49PAyQiIiL/YL4nIiIKHpIQQpzrQdasWYPVq1ejbdu2GDZsmC/iOu9w7V8iIgo0vs5NF3q+Z64nIqJAU5PcVKul/s7Wt29f9O3b1xeHIiIiogDFfE9ERHT+qlW3/+nTp+PTTz+t8Pmnn36K//3vf+ccFBEREfkf8z0REVHwqFXx/9FHH6FDhw4VPu/UqRM+/PDDcw6KiIiI/I/5noiIKHjUeqm/pKSkCp83aNDANSswERERnd+Y74mIiIJHrYr/pk2bYtWqVRU+X7VqFRo1anTOQREREZH/Md8TEREFj1pN+Hfvvfdi4sSJsNvtuOqqqwAAaWlpmDJlCh577DGfBkhERET+wXxPREQUPGpV/E+ePBknT57Egw8+CJvNBgAICQnB448/jqlTp/o0QCIiIvIP5nsiIqLgIQkhRG13Lioqws6dOxEaGoq2bdvCaDT6MrbzCtf+JSKiQOOr3MR878RcT0REgaYmualWT/7LRURE4JJLLjmXQxAREVGAY74nIiI6/1W7+B8xYgRmzZoFk8mEESNGeG37008/nXNgREREVP+Y74mIiIJTtYv/qKgoSJLk+jMREREFH+Z7IiKi4HROY/7pNI4DJCKiQMPc5Fv8PomIKNDUJDfJ9RQTEREREREREflJtbv9X3zxxa5ugFXZuHFjrQMiIiIi/2G+JyIiCk7VLv6HDx9eh2EQERFRIGC+JyIiCk4c8+8jHAdIRESBhrnJt/h9EhFRoKmXMf8FBQWYOXMmpk6diry8PADO7n9Hjhyp7SGJiIgowDDfExERBYdqd/s/05YtW5CSkoKoqCgcOHAA9957L2JjY/HTTz8hKysLX3zxha/jJCIionrGfE9ERBQ8avXkf9KkSbjrrrvwzz//ICQkxPX5kCFDsGLFCp8FR0RERP7DfE9ERBQ8alX8r1u3Dvfff3+Fzxs3bozs7OxqH2f69Om45JJLEBkZiYYNG2L48OHYvXu3W5uysjKMHz8ecXFxiIiIwMiRI5GTk+PWJisrC0OHDkVYWBgaNmyIyZMnw+FwuLVZtmwZevToAaPRiDZt2mDWrFkV4pkxYwZatGiBkJAQ9OnTB2vXrq32tRAREQUbX+R75noiIqLAUKvi32g0wmKxVPh8z549aNCgQbWPs3z5cowfPx5r1qzB4sWLYbfbcfXVV6O4uNjV5tFHH8W8efMwZ84cLF++HEePHsWIESNc21VVxdChQ2Gz2bB69Wp8/vnnmDVrFp5++mlXm8zMTAwdOhRXXnklMjIyMHHiRIwdOxaLFi1ytZk9ezYmTZqEZ555Bhs3bkS3bt2QmpqK3Nzcmn49REREQcEX+Z65noiIKECIWhgzZowYPny4sNlsIiIiQuzfv18cPHhQXHzxxeKRRx6pzSGFEELk5uYKAGL58uVCCCEKCgqEXq8Xc+bMcbXZuXOnACDS09OFEEIsWLBAyLIssrOzXW0++OADYTKZhNVqFUIIMWXKFNGpUye3c40aNUqkpqa63vfu3VuMHz/e9V5VVdGoUSMxffr0asVuNpsFAGE2m2t41URERHXjXHNTXeR75noiIiLfqUluqtWT/9dffx1FRUVo0KABSktLMWDAALRp0waRkZF48cUXa30jwmw2AwBiY2MBABs2bIDdbkdKSoqrTYcOHdCsWTOkp6cDANLT09GlSxckJCS42qSmpsJisWD79u2uNmceo7xN+TFsNhs2bNjg1kaWZaSkpLjanM1qtcJisbi9iIiIgkld5HvmeiIiIv+o1Wz/UVFRWLx4MVatWoXNmzejqKgIPXr0qJB0a0LTNEycOBH9+vVD586dAQDZ2dkwGAyIjo52a5uQkOAaa5idne32Y6B8e/k2b20sFgtKS0uRn58PVVUrbbNr165K450+fTqmTZtWu4slIiI6D/g63zPXExER+U+Ni39N0zBr1iz89NNPOHDgACRJQsuWLZGYmAghBCRJqlUg48ePx7Zt27By5cpa7V/fpk6dikmTJrneWywWNG3a1I8RERER+U5d5HvmeiIiIv+pUbd/IQSuu+46jB07FkeOHEGXLl3QqVMnHDx4EHfddRduuOGGWgUxYcIE/Pbbb1i6dCmaNGni+jwxMRE2mw0FBQVu7XNycpCYmOhqc/aMwOXvq2pjMpkQGhqK+Ph4KIpSaZvyY5zNaDTCZDK5vYiIiIJBXeR75noiIiL/qlHxP2vWLKxYsQJpaWnYtGkTvv32W3z33XfYvHkz/vzzTyxZsgRffPFFtY8nhMCECRPw888/Y8mSJWjZsqXb9p49e0Kv1yMtLc312e7du5GVlYXk5GQAQHJyMrZu3eo2U+/ixYthMpnQsWNHV5szj1HepvwYBoMBPXv2dGujaRrS0tJcbYiIiC4Uvsz3zPVEREQBoiYzCQ4aNMjrjLgvvviiuPrqq6t9vAceeEBERUWJZcuWiWPHjrleJSUlrjbjxo0TzZo1E0uWLBHr168XycnJIjk52bXd4XCIzp07i6uvvlpkZGSIhQsXigYNGoipU6e62uzfv1+EhYWJyZMni507d4oZM2YIRVHEwoULXW2+++47YTQaxaxZs8SOHTvEfffdJ6Kjo91mFvaGMwATEVGgqW1u8mW+Z64nIiKqOzXJTTUq/hMSEsSmTZs8bt+4caNISEio9vEAVPr67LPPXG1KS0vFgw8+KGJiYkRYWJi44YYbxLFjx9yOc+DAAXHNNdeI0NBQER8fLx577DFht9vd2ixdulR0795dGAwG0apVK7dzlHv33XdFs2bNhMFgEL179xZr1qyp9rXwBwEREQWa2uYmX+Z75noiIqK6U5PcJAkhRHV7CRgMBhw8eBBJSUmVbj969ChatmwJq9Vaq14I5zOLxYKoqCiYzWaOCSQiooBQ29zEfF855noiIgo0NclNNRrzr6oqdDrPCwQoigKHw1GTQxIREVGAYb4nIiIKPjVa6k8IgbvuugtGo7HS7RfaEwAiIqJgxHxPREQUfGpU/I8ePbrKNnfeeWetgyEiIiL/Y74nIiIKPjUq/j/77LO6ioOIiIgCBPM9ERFR8KnRmH8iIiIiIiIiOv+w+CciIiIiIiIKciz+iYiIiIiIiIIci38iIiIiIiKiIMfin4iIiIiIiCjIsfgnIiIiIiIiCnIs/omIiIiIiIiCHIt/IiIiIiIioiDH4p+IiIiIiIgoyOn8HQAFnvzCEhzMzUeIQY+2jeOhyLxHREREFEyspVbs35IFIQRadm6K0IhQf4dERER1jMU/uRwvKMJrPyxH2qZ/oGkCAJAQE4H7hvTFDZd18XN0REREdK4cdge+nDYHc99biBJLCQAgJNyIofem4O4Xb4Ux1OjnCImIqK6w+CcAwElLMe585TucMBe5Cn8AyMkvwvNf/4m8wlKMuaa3HyMkIiKic6FpGl4Y9SZW/7IOQpzO9WXFVvz8zgLszTiAlxf9Fzo9fx4SEQUj9ucmAMD//b4WJ8xFUM8o/M/0wbzVyMkvrOeoiIiIyFfWLtiEVXPXuhX+5TRNYPOy7Vj67So/REZERPWBxT/Brqr4ZfV2j4V/uXlrdtRTRERERORr8z9ZDFnx/NNPliXM/3hxPUZERET1icU/obDYilKb3WsbSZJw9ISlniIiIiIiXzu6LweaqnncrmkCR/fn1GNERERUn1j8E8JCDJBlqcp2UREh9RANERER1YWo+EhIkvd8HxUXWU/REBFRfWPxTwgx6HBltzZQvNwAUDUNg3u1r8eoiIiIyJdS/jWg0vH+5SRZwqA7B9RjREREVJ9Y/BMA4N4hfaDIMuRKngjIkoSUHm3RvmlDP0RGREREvjDw9svQtH0jKLqKP/9knYz4xrG4ZuxAP0RGRET1gcU/AQDaNWmA9x8egQZR4QCck/5IEiBJwNA+F+H5uwb7OUIiIiI6F8ZQI15b+iw6Jjt78kmy5Br217pbC7yx/DlExkT4M0QiIqpDkvDW/4uqzWKxICoqCmazGSaTyd/h1JqqaViz4yD2HTuJEIMel3dphcRYjv8jIjofBUtuChTB9H3uzcjE5qXbIYRA58s6oP0lbaqcD4CIiAJPTXKTrp5iovOEIsvo17kl+nVu6e9QiIiIqI606d4Sbboz1xMRXUjY7Z+IiIiIiIgoyLH4JyIiIiIiIgpyLP6JiIiIiIiIghzH/FOlbFY79qzfB4fNgRadmyK6QZS/QyIiIiIfEkJg/5aDMJ8oRMNm8WjSNsnfIRERUR1i8U9uNE3Ddy/PxQ+v/4rC/GIAgKKTccWofnjgzbtgtzmweel2aKqGi/q2RZN2jfwcMREREdXUmt824OMpX+LQriOuzzpe2h4PvnU3mndsgg1/bEZhfjEat0lE58s6cCUAIqIgwKX+fCRYlv95c9xHWPDxnxU+lxQJoWEhKC0ug9BO/5XpkdIFk2dNQHyj2PoMk4iIqiFYclOgCJbvc/mcdLxwyxuQIOHMn4GSLEGWJegMelhLrK7PG7VJxKSPx6HbFZ38ES4REXlRk9zEMf/ksmfDvkoLfwAQqkBJYalb4Q8AGcu249H+T6GooLg+QiQiIqJzYLfZ8e74TwABnP38R2gCqkNzK/wB4Nj+HDyR+jx2rNlTn6ESEZGPsfgnl4WfLoWiU2q0j+bQkHPwOH77aHEdRUVERES+su73DJhPFNZoH6EJaJrAp09+U0dRERFRfWDxTy65B49Ddag13k9oAgs/XVIHEREREZEv5Rw8Dkmu+fh9TdWwedl2HD98sg6iIiKi+sDin1xMDSKh6Gr3VyI/p8C3wRAREZHPRcVHVhjCVxPM90RE5y8W/+SScvvlUB1azXeUgPjGnPCPiIgo0PUd1gvGUEPtdpaAOE7wS0R03mLxTy7dr+qM7ld2hqzU7K+FBAlDxqbUUVRERETkK2GRobjjmZtrvJ+syOg1qBvikmLqICoiIqoPLP7JRZZlPPfLFPQf2Rc4NRywfF3fkIiQSscIyoqMph0aYci9A+szVCIiIqqlmydfh7Ev/8vVA6A8vxvDKu8RICsydAYdxrx8e73FSEREvieJs9d5oVoJlrV/y+UcPI71izJgtzrQtmcrtLm4BT6Z8hUWzPwTdqsDgPPHQv8RffDIB/fBFBfp54iJiOhswZab/C3Yvs+SwlKk/7oe5hMWJDRvgN5DLsYfs5Zh1tOzUZBrdrVr26MVHvnwPrTv1dqP0RIRUWVqkptY/PtIsP0g8KQwvwg70vdAdaho16s14jn2j4goYF0ouam+XCjfp8PuwPZVu1FUUIxGrRPQsktzf4dEREQe1CQ36eopJgoSkTER6DOkh7/DICIiojqi0+vQ7YpO/g6DiIh8jGP+iYiIiIiIiIIci38iIiIiIiKiIMfin4iIiIiIiCjIsfgnIiIiIiIiCnIs/omIiIiIiIiCHIt/IiIiIiIioiDH4p+IiIiIiIgoyLH4JyIiIiIiIgpyLP6JiIiIiIiIghyLfyIiIiIiIqIgx+KfiIiIiIiIKMix+CciIiIiIiIKciz+iYiIiIiIiIIci38iIiIiIiKiIMfin4iIiIiIiCjIsfgnIiIiIiIiCnIs/omIiIiIiIiCHIt/IiIiIiIioiDH4p+IiIiIiIgoyLH4JyIiIiIiIgpyLP6JiIiIiIiIgpxfi/8VK1Zg2LBhaNSoESRJwty5c922CyHw9NNPIykpCaGhoUhJScE///zj1iYvLw+33347TCYToqOjMWbMGBQVFbm12bJlC/r374+QkBA0bdoUr7zySoVY5syZgw4dOiAkJARdunTBggULfH69REREFxrmeiIiosDg1+K/uLgY3bp1w4wZMyrd/sorr+Cdd97Bhx9+iL///hvh4eFITU1FWVmZq83tt9+O7du3Y/Hixfjtt9+wYsUK3Hfffa7tFosFV199NZo3b44NGzbg1VdfxbPPPouPP/7Y1Wb16tW49dZbMWbMGGzatAnDhw/H8OHDsW3btrq7eCIiogsAcz0REVGAEAECgPj5559d7zVNE4mJieLVV191fVZQUCCMRqP49ttvhRBC7NixQwAQ69atc7X5/fffhSRJ4siRI0IIId5//30RExMjrFarq83jjz8u2rdv73p/8803i6FDh7rF06dPH3H//fdXO36z2SwACLPZXO19iIiI6lKg5SbmeiIiIt+qSW4K2DH/mZmZyM7ORkpKiuuzqKgo9OnTB+np6QCA9PR0REdHo1evXq42KSkpkGUZf//9t6vN5ZdfDoPB4GqTmpqK3bt3Iz8/39XmzPOUtyk/T2WsVissFovbi4iIiKqPuZ6IiKj+BGzxn52dDQBISEhw+zwhIcG1LTs7Gw0bNnTbrtPpEBsb69amsmOceQ5Pbcq3V2b69OmIiopyvZo2bVrTSyQiIrqgMdcTERHVn4At/gPd1KlTYTabXa9Dhw75OyQiIiLyIeZ6IiIKJgFb/CcmJgIAcnJy3D7PyclxbUtMTERubq7bdofDgby8PLc2lR3jzHN4alO+vTJGoxEmk8ntRURERNXHXE9ERFR/Arb4b9myJRITE5GWlub6zGKx4O+//0ZycjIAIDk5GQUFBdiwYYOrzZIlS6BpGvr06eNqs2LFCtjtdlebxYsXo3379oiJiXG1OfM85W3Kz0NERES+x1xPRERUf/xa/BcVFSEjIwMZGRkAnBP/ZGRkICsrC5IkYeLEiXjhhRfw66+/YuvWrbjzzjvRqFEjDB8+HABw0UUXYfDgwbj33nuxdu1arFq1ChMmTMAtt9yCRo0aAQBuu+02GAwGjBkzBtu3b8fs2bPx9ttvY9KkSa44HnnkESxcuBCvv/46du3ahWeffRbr16/HhAkT6vsrISIiCirM9URERAGiHlYf8Gjp0qUCQIXX6NGjhRDOJYCeeuopkZCQIIxGoxg4cKDYvXu32zFOnjwpbr31VhERESFMJpO4++67RWFhoVubzZs3i8suu0wYjUbRuHFj8fLLL1eI5fvvvxft2rUTBoNBdOrUScyfP79G18Llfy48docqHKrq7zCIiDwKhNzEXE/nM1VVhc1q83cYREQe1SQ3SUII4Y+bDsHGYrEgKioKZrOZYwKDmBACf2zYg6/+3IDtB3MgAejWuhHuHNQLV3Rr7e/wiIjcMDf5Fr/PC8fu9fvw/StzseqXdVDtKhJbNMR14wfj+gmDYTDq/R0eEZFLTXITi38f4Q+C4CeEwJs/rsBXaRshSxK0U//XKf/z/UP74v5rOXaUiAIHc5Nv8fu8MKz+dR2eu/E1AIDq0FyfS7KEzpd1wMsL/wtDiMFf4RERualJbgrYCf+IAs3aXVn4Km0jALgK/zP//NH8NdiaecwvsREREdG5K7aUYPptb0NTNbfCHwCEJrBt5S7MeX2en6IjIjo3LP6Jqmn2ss1QZMnjdkWWMGf55nqMiIiIiHxpyTcrUVZqhad+sUIT+GXGQqiqWr+BERH5AIt/omrakZUDVfM8SkbVBHYczPG4nYiIiALbvowDUHSK1zb52QWwnCyqp4iIiHxH5+8AyDeK7GX45fB6LDi6EQW2EjQOi8XwJpfg6qSu0MnekxhVj0Ff9fdoNPD/UkREVDdUVcXKn9bit4/+wKFdRxERHYarbuuPofelICqecxD4gt6oc65HUQWDkfmeiM4/fPIfBHJKC3D76nfwzu4F2FuYjeNWC7bkH8SzW+fgofWfoky1+zvEoHBV9zaQvXT7lyUJV3ZvU48RERHRhcJhd+C5G1/HC6PewJblO3DyaB4O7jiMWU9/h3u7TMKh3Uf8HWJQSB7WC6rDc5d+WZHRqV97hEeF12NURES+weI/CDyZ8S1yyyyuxZMBQDv1p015B/D+nj/8FlswuXlAd+gVBbJU8QaALEkIC9Hjhn6d/RAZEREFu+9enov0X9cDADT19ER0QhMwnyjE08NfgaZpnnanaup+VWe07t4Ciq7yn8iaquHWqSPqOSoiIt9g8X+e22U+gq3mQ1BF5Qlfg8DcQ2tR4rDWc2TBp1GcCe9OGI4Qox4SnAV/+Y2AyDAj3n94JOJMfBJARES+5bA7MPfdBfC0OrOmaji8+ygylmyr58iCjyzLeHH+k2jaoTEAQNHJkGQJsiJDliU89N5Y9BnSw89REhHVDgcsnecy8g9AggThZYBamWbHP4XZ6BbTvB4jC0692jXFwpfGYsHaXdj4z2FIkoRL2jfFNZd0QKhR7+/wiIgoCB3Zmw3ziUKvbRSdgq1/7USPlK71FFXwikuKwYebXsXaBZvw109rUFZsRctOzTB4zFVo0CTO3+EREdUai//znATPY9Dd25GvRIQacfOAbrh5QDd/h0JERBcAqZLhZhWJaraj6lAUBcnDeiF5WC9/h0JE5DPs9n+e6xnXyutTfwAIVQxoZ0qqp4iIiIjIlxq3SURMQpTXNqpDQ7crO9VTREREdD7ik//zXJvIRPSIbYnN+QcrHfcvQcKNzfoiRDFUeSwhVMC6AsL6JyCskPQdgNARkOTYugidiIiIqkHRKRj56DDMnPpVpcvQyToZzS9qgq6Xd6zW8fZmZOLPL5YjP9eM+EaxGDT6CrTo1NTHURMRUaBh8R8EXuh2Cx5Y+wkOFp9wjf+XIUGDwKUN2mFc25QqjyHUbIj8MYDjHzj/WgiIsnlA4ZtA1HRIodfV+XUQERFR5W587Frs33IAS75ZCVmRnTP+n+rlH5cUg2lzp1TZ7d9hd+D1MR/gz69WQNEpEEJAkoDvX/sV194/CA/NGAtZZqdQIqJgxeI/CMQbI/HlpROw8OhmLDi6Cfm2IjQOjcXwpr3Rv2EHKJL3RC6ECpF3D6BmnvrEccZWO4R5MqAkQTJcUmfXQERERJ4pioInvnwYKf+6HL99tBiHdh1BREwErrr1MgwaPQDhprAqjzHzia+R9vVfAFBhLfvfPl6M2MQY3PHMTXUSPxER+Z8kPK0bQzVisVgQFRUFs9kMk8lUp+cy20pQ7LAizhgBo3LuM8yLsqUQBfd7aaEAhv6QYz+u8lilNjsWrt2FBWt3wVxcimYJMRh5WVf0vagZJyIiIqpn9ZmbLgT1+X2WlVhRkGtGRHQ4IqLPfRnZwvwi3Jx0Lxw2h8c2oZEh+P7YTISEGb0eSwiB9X9sxoJP/sThPUcRFW/CwNv748pbL6tyXyIi8q2a5CY++Q9wQgjsK8pBkaMM+dZizMlag/V5+wAARlmPYU16YGzrgYg1RtT+HNbFABQAqocWKmBbDiHskCTPNxuOFxTh3jfnICu3AJIECAFkZudhyaa9uLpnO7xw9zXQKexOSEREdLbcrOPIzToBu82BtK//QtrXfzkLdQnoM6QH7njmZrTv1brWx9+4eIvXwh8ASgvLsPWvnbgktbvHNqpDxfR/vYPl36+GrJOhOTRIkoTNy7Zj9iu/4LWlzyK+EecKIiIKRCz+A9jiY1vw4T+LcajkZKXbrZodPx9ah1XHd+PTvg8gzhhZuxMJKyqdQci9EZzDATwX///++DccOWF2tj51OFVz/mHxhj1o0ygeY4f0qV2MREREQWj/loP4cNIsbFqyrfIGAli3MAMbFm/Bywv/i25X1G5Gf2uprVrtbFW0+3b6z1gxZzUAQHM4Jxou70SanZmD5296HW+verFWMRIRUd3iY9gA9fOhtfjP5u88Fv7lVKEht8yCD/b8UetzSboOVTeSGwEI8bh5x8FsbM085ir2zyYAfL1kI+wOT70LiIiILiz7txzEI/3+g83Ld3htp6kaVIeKl+98F6pauzzaqlvzarVr0dnzrP92mx0/vT0fngaMqg4NO9L3YM+GfbUJkYiI6hiL/wBUZC/DGzt/q3Z7VWj4/VgGih3W2p0wbCSc3f49kSCF3+F1zP7a3Ycgy97H9JuLy7D/mPvNDMvJQiz+cjnmfbAIGUu3QdMqLldIREQUjGY88ilsZXbnzP1VEJrAicMnsXHxllqdq033lmjXqzVkD8PvZEVG96s6o3GbJI/HOLj9MArziryeR1ZkZJzVi8Fus2P1r+vw6/uLsOKHdJSV1PL3ChERnRN2+w9Ai7O3wKZ5H5d3NrumIru0AK0jEyCEwLqT+zD38DpkFZ9AtCEc1zTqjpTELpVOECjJsUDUixDmx+G8H3TmUwUJMPQBwu7wen4hBKoznZ926nGB6lAx84mvMPfd3+Gwq645ApJaJeDxLx5Cp0vbV/vaiYiIzjfHMnOwpYon/meTFRlZO4/gksEXAwCydh3BvA8WYUf6Hih6BX2G9MCQsQMRkxBd6f5TPp+ARy/7L0oKS6E6NLfjRsVHYtIn47yev7o36LUzegEu+XYlZjz8KSwnC125PswUintevA3Xjx9creMREZFvsPgPQEdL8qFIMhyiZk/Bw3QGODQVz2z5Houzt0KRZKhCgwQJa0/uxReZK/D+JWMqnRtACh0OKI0hij4CbH8BEICcBCn8TiDsDkiSweu5u7Zq5LHLf7nwEANaJsYBAN596P+w4OPFrq6D5f/MOZCLySnT8M7qF9Gme8saXT8R0flMaBag7DcIxwFIsgkIuQaSrvYTvFFgyzlwvMb7aKqG0AjnELzfPlqMdx78BJIiucbe7/r7H8z+31y8OP9JdOl/UYX9m1/UBO9veAWz/zcXi79YDmupDaERIUi960qMevx6xDeO83r+5h2bIDQyBKWFZV5j7NzPeQP/rx/XYPrtb7u2lef6Eksp3nvo/yBJEq57MLVG3wER0flMdahIn7ceO1bvhqzI6JHSFRcP7FJvq6JxqT8f8eXyP19mrsCM3YugVTkJn5MECW0jE/FVv4cwc28aPtmbVumeiiSjW0xzfNj7Xq/HE8IGCAcghVb7L6IQAje/8CUOZOdVehNAliTcOagnHr6hP47uy8botg95PJasyOh7bU9M+3lKtc5NRHS+EyU/QFimAbDBOQxLAFCBkKGQol6GJNVu+TQu9edbvvw+920+gHEXT67RPrJOxrdZH+Lovhw8evlTlc7VK8sSjOFGfLX/fZjiPE8ErKoqrCU2hIQbIcvVHwX6yZQvMeeNeRCV5XqdjOYXNcFHGa9BCIHRbR9Cdmaux2OFR4Xh+2OfwBDi/QEDEVEw+Gfjfjw9/BWcOHwSil4BhPNmQPNOTfH8r48jqWVCrY5bk9zEMf8BKCWxSzXLficBgXvbDIRNc+C7A6s97qsKDRvzMrHHcszr8STJAEkOq9EdKEmS8Pr9wxAdEQr5jP3K/9yrfVPcf20yAGDJNys9jjkEnE8N0uetR7G5uNrnJyI6X4myNAjLkwDKV15xwDX8qux3CPN//Rcc1ZlWXZujSbskVDfVSpKEa+8fhNjEGPz45m9QPORRTRMoK7Zi0WdLvR5PURSERYbWqPAHgNHPjXKtOCCdMdePJEuIbhCFZ378NyRJwq61e70W/gBQbC7BuoUZNTo/EdH56MSRk5icMg15x/IBAKpdhXpqIvTDu49g8lXTUFrsuVeVr7DbfwBKCo3BiKa98dOhvz0W8hIkAAI6WcHki67DgISO2F5wCIUnCiFpgIhWgFNJWQgADglQBBRZwq+H1yMpNAaKJOGSuNZoHZlY61iFEK6bBM0axuD7p+7ET39txfy/d8BSYkXTBtEYeXkXpPZqD73inFTQfNwCWZageZmwWGgChfnFCI8Kr3VsRESBTggBUfQWcOq/6RVpQNmvEI6HIOma1W9wVKckScLYl/+FZ0e86qWR8ya6pglcfdcVeOCNuwAAG/7Y7DZm/2xCE1g9bz0MoQbYSm1o2bU5eqR0qXGh7zreqU6ikiTBEGLA9N//g6XfrsL8jxfj6L5sRMREYNAdAzD0vhRXbwPzcUu1jl3ddkRE57NfZixCaWFZpRO8qg4NOVnHseTrvzD0vkF1GgeL/wD12EXXQi8rmJO15tS4/bN/FgoIAIMSumBoo4ux8NMl+PJ/P8L0j/Muu9ZQh7JhMShr1wjiUCiUoyUI3XwEhqw8LFT3QGtphO36aNiuisQl8W3wQvdRiDFEVCs2VdMwL30HvluWgb1HTkCnyBjQrTVGD+qJjs0TMeaa3hhzTW+P+zdoGge1ipmNdQYdohqwiyoRBTn1MODYXUUjCbD+AejG1ktIVH/6De+NqV89jHfGz0SxuQSSLFXoTq9pAs07NsE9L96KXWv34uvnf0BpUdVPh7b9tRPb/toJSZIghEBC8wZ48tuJ6Ni3XbXj27x8O354fR7W/7EZmqqh/SWtMeKRoRhw86UYdOcADLpzgMd9GzaLr9Y5GlSzHRHR+Wzptyu9ruwiQcKy2avrvPjnmH8fqatxlSethZi+fS5W5O702Cbk4+Mwzi1we3BU/i/V1iYOtuaxiFjyDwBAKt8uOf9sGxAB25RGaBHZEJ9fOh4G2fv9IFXT8MTM+UjbtNc1ay8AKLIEIYCXxw5BSg/vPyxOHsvHbc3Gefw/gKKTMfD2yzH5s/Fej0NEdL4T9l0QJ6+ropUOCL8fcuQjNT4+x/z7Vl19n7YyG75/9Vd8/uzsyjuAnKG8mK8p51N7PWas/x+aX9SkyvbzPvwD7zz4CWSd7JpQUJadvRCuezAVE94d43V4oBAC4y6ejMxtWZXODwAJiE2MxjcHP4Si87bcMBHR+W9E/N1VLpXatkcrvL/+fzU+Nsf8B5EwxYCNCzcj7NkjiBydiYhxB2H85iSkfOdSgMr2UmfhD7j9YJBOvYx7TzoLf3G68AdO/1m/vAjyHwXYV5SDtOytVcbz86ptSNu013m6M46nagJCCPzn09+RX1Tq9RhxSTG489mbK90mKzLCTGG445mbqoyFiOi8pzRC1Z3wHJB0XP0kmBlCDNi0pOocDKBWhX/5fg67A99O/6nKtof/OYZ3x88EAFfhD5xewu/X9xdh9S/rvB5DkiRMeHcMZFl2mxugfBsAPPTeWBb+RHRBaNahMWTZ8w1TRSejeaeqb8yeKxb/AUzTNDxz15tQ/nsQuvUlkI87oGTZYPwmDxH3H4S83wrDAjOEl7wpTv2Px79qEmD8pQAyJPx+NKPKmL5dssnjsQQAhybw6+rtVR7ntidH4OEZYxGTGO32+cVXdca7a15CYouGVR6DiOh851zSbyicM/xX2gKQIoGQq+szLKpnf3y+FFuW76jyqf+5Uh0als9e7ZpkypPfPvyjQsF+JlmRMffd36s8X5f+F+HVtGfQulsLt88bt0vCC78+gctu6FOtuImIznfDHkh13UCtjOrQ6rzLP8Ax/wFt/keLsembDQAA6Ywe8pIGoERD+DNHICIVSF5yuKcppFzbBaAcsEFTNeRbvc+ub3eoyMzO89pGArAzK8drG8B513/YA6kYcm8Kdv79D0qLytC0fSMW/UR0wZEi/w1h+xvQjsM1yz+A8vvzUtR0SFKIX2Kjundw52G8NuaDejufw66irLjM64S6u9ft9To2VVM17Fm/r1rn69L/Inyw4RVkbsvCicMnEd0wCm0ubllva1oTEQWCK265FMu/X401v21w78F1qli7fsJgdO7Xoc7jYPEfQMy2EvyQtQa/Hl6PfGsRQl7Y77GtpAHSSRWa3Tnx37mkUCEDiiKjaXic13ayLLmN8680LgnQ16AL34kjeVAdKuIbxSCheYNq70dEFCwkJQGI+xGi6G2gdC4Am3ODvhekiIcgGfl0NJioqoo/v1yBX99fhKydh51Fdj1OvxQeFYbQyFCvbfRGfZXH0Rmqn+uLCopRXFCM8OhwNO/UlIU/EV1wFEXB0z88hjmvzcPP78xHfo4ZAJDUMgE3T74eQ+9LqZc4WPwHiJwyM+5d8xFyy8zQICDlO2A8ZvW6jwAgWTSvhX9VNwaEDDh6hkGFwPVNenk9nyLLSL6oOf7elQXVQ7cVVRO4rHPVY1MPbD+E9yd+hk1pp8c4NmnfCGNeuo3dAInogiMpDSBFvQAR+SSgnQDkCEhyrL/DIh9THSqm3fga0n9dX+nM/nVNkiUMvTelyiX/+g7tiYwl2zzOL6DoZCRfd0mV5yu2lODjyV9i8RfLYLc65yqKiA7HyEevxa1P3gBF4Xh/Irpw6PQ63Dr1Btw8+TrkHjoBRZHRoGl8vd4Q5Zj/ADFtyxwct1qglXfSr8bfgfJJ/YDKu/YLCRB6BWqoHqKS4wkA0ADbjTFISeyC3nFtPJ7LbrNj2exVSNyWA8Pag1COV5ytUpElJMVG4qruno8DAAd3HMLDlz6Jzcvc5wY4sucopo18DYu/XO51fyKiYCXJYZB0zVj4B6mf3pqPNfOcw/nqu/CH5Fx+7+Yp13tsIoTAtpU7ceJoHvRGXeXj/k99NOKRoV5PV1ZixeSrpmHhp0tchT/g7AXwxbOz8epdM2o9eSER0flM0SlIapmAhs0a1HtPKD75DwAHi45jfZ57F38RpUBtpId81F6tLv2umwAyAM35PiQqDLlXtYFkUBAxbzukYtvpmwQyAAGIiY1x1/BrcU/rKz3+5ctYug0vjHoD5hOFUPQKwjWBsA2HYW9kQuGg9pBCDdCEQIOoCHzwyMgqu/1/NPlLWEtsFcYTlv8GeO+h/0P/kX0REmasxpUTEREFPk3T8NM7C3xa8EqSBFmRIMkyHDaHl4bAFaP64cE370JUfOXLQJ04moenr/8f/tmwH4rO+Wzo7BsUkixB0Sn4z7cT0aprc6+x/T4zDXs3ZVZ6vUIAaV//haH3DUKX/hdVcZVEROQrLP4DwA7L4YofShKsI2MQ9m5utY8jJABGCaGKAY998gD63dAbuZYS/Jq+HVkDuqB082HImSdhkIGYjg1w6ehk9L7oIhgVz2P7Mrdl4ckhL8Jhd05CpdpPT0ZlyC5E42X70eGJobi8e2sM6tEORr33v1InjuZh3cJNXmchLLGUYtXPazHw9v7VvnYiIqJAZj5uwYnDJ316zFueGI4bJw2DoleQ9tVf2Pn3HphzzbCW2aE36hAaHoJeqd3R74beiG4Q5fE4NqsdkwdOw9F92QCcs06Xc95gkNFtQEd0u7IzrhlzFWISoquM7bePFkN4SfaKTsbv/5fG4p+IqB6x+A8AilT56Av7YBOs+8pgXGCBkN1n/K+MJACUCtz87PW4YlQ/AEDj+Cg8MOzSCm2FVgKIEsDLUj4A8P2rv0BTtUq7JwpNoPTgSdzUoiku6dPRe3CnHD90ssqljGITNBTnbYfQekCSPc9GTEREdL7w9Xr2HXq3wT0v3uZ6f92DqbjuwVS3NnabHUX5xQiN8L5axF8/rMHh3Ucr3SaEgBACLbs2x21Pjqh2fDkHj3vN96pDQ+a2LBQcN3u9MUFERL7DMf8BoEdsK8iVde6XJJSNb4iilxrD0TccQlf1EsAhEUZcP2Gwx+3ClgEt7z6I3Ishjl8KkdsbmuVlCM1csa0QWP59utsTgLMpOhnLZ6+uIqrTTHERHrd16VuEV3/ci283bcW1N73hjK3gCQg1u5LYbBBaEccLEhHReSEyNgKtuzWvfBx9LYx74y6P204czcM7D36CG2Luws1J9+I60514ftTryNyWVWn7Zd+v8hqXpmpY8s3KGsVnivWc78vt3ZiJmxLGYvLAadi2aleF7aqqothSAlX1sqYxERFVG4v/ABBvjMQ1jS/2eANA7R6Gkv82Qtm9VS+FN/WrR2CKjax0myhbCpF3K2D7C67bCKIIKPkc4uRNEFq+W3vVocJutXs9n6YKlBSWVBlXucZtkpzr+571IyN5sBmvzNmHTr2Lz/jUDpT9AnFypOsGgLCtg5Y3FiKnC0RuD4jjAyCKPoYQtmrHQEREVN8kScKox2/wyUR/V991BTpd2r7SbblZxzH+kicwf+afsJY6c6Omalj181pM6D0V21fvrrBPUX5xlXGVFpXWKMZBdw6ArFTvZ+aWFTvw7yufwdrfN7mu4e0HPsb1pjsxPHo0ro8ajXfGz0TuoRM1ioGIiNyx+A8Qj3e8Dj1iWwHA6ZsAZ+VhW6oJanujc1K/s0lA6t1XInlY5cv1CVEGYf43AA3A2XfQVUDNgsh/AKLkOwj1GADnchTxTbzPOC3JAo3bRHttc7Yx0293W4NQb9Tw2BuHAAmouOqPCmh5EIWvQJT+BpH3L8C2Cq4vR8uGKHodIu9uCOF9aUQiIiJ/uvKWfrjzmZsBoNqF8ZlkRUaT9o3wgJen/u89/CkKcs3Qzuq1pzo02G12PH39y/jprfnYtfYfV++55hc1cU3y50l8k7gaxXr9hMEwxUVW6zo1VYOmCrxy13s4sC0L43pMwYL/S3PdvLCWWLHgk8V4sOcUHNl7rEZxEBHRaSz+A0SIYsArHW5Fr98jgPRCQBOoUP0bZBS/1AS2oVEQhtNPzpVoAxqP74gNo224eeVbeG7rD9hhPj2J4OE9R/HnzP8AorDiMV00wL4RwvI0xPEroBVMgRBluO6Bwd67AmoCg2/4BMK+rdrX2uvqbnj6h8dcPRQuG1KIyGgVnpcdVoGy3yHMj5+K/+ybFwKwrweKZ1U7BiIiIn/419M34vrxg6FpVUzkczYJSGgWj9CIEEzoOxVTr3kRf/30t6tLvCWvELOeno30X9dXWE2nnNAELCeL8MFjs/BQ3ycxvvcTOLovG0PuS/E6xA8ADu8+it8+WlztcGMTY/DmiufQolNTV/zeCCFgPm7BtBtfQ7G5pNKbF4X5xXh97AfVjoGIiNxJgoOmfcJisSAqKgpmsxkmU+XL6HgjhMCUQc9h4+G9KJzeCAitYmKgEg3KIRuEAmgtjIDujJsBkgxVaJjQbjD0a4rw/vJ5GDsiE7d0zoReqe6/bhkwDsCS+bfhzfs+gs1qO7WcgJMkCQgh4a4njuHWh08AcgykBsshSYZqX7PdZsff8zciKuwrXNTlT8hyVWP6JHid9UBuCKnBX/W+XiYRUaA619xE7nzxfc557Vd8POXLc45FVmRoqoZeqd1x85Tr8fR1L6OsuGY94GRFRnTDKPznm0fwyt0zkHPguPcdJOC9v19G+16tq30OIQR2rtmDZbNX4+d3FlQRjwRNrfp3yv/teAvNOjSudgxERMGsJrmJs/0HiC0rdmDTX9tR9HkLwFiNDhlhMtT2Z83ea9OgX1UE/coioEzDzITPYRseDW1YNEoahsLDogIeaIB1KX556xBsZeEYdFMeVv0ejZIi502Jpm2suPWRHFw1ouBU85NA2UIg9Lpqn0Fv0OOyG/pAFO+AVlidpwlVFP9arnMOA6nyOQ/8Sdh3Q5T9CmhmSEpjIPQGSEqiv8MiIqJ6VFpchi+fm+OTY5U/3V+/KAPrF2XU+hj5OQV4/OrnoXroLXAmRVEw990FePzzh6p9DkmS0DG5PcJMYVUW/9Up/AEgc2tWQBb/+TkF+OPz5Ti69xgiosMxYNSlaNez+jdKiIjqGov/ALH021VQrzBBROsAVQBCADV4gi3l2BE+9QiUbDuE5Fz2T8ilMC60oGx0HJaPboT7k7bXKCaHQ8LAkfnI2hOCf799GA//7whyDhtgCNHQsLH9rPB0ELa1kKoo/vNzzchYsg0OmwPterVC845NkZvbA/FGzeMgFAEJkmQCRDGccxZ4/BaAGvQ8qA9C2CDMU4CyBQAUABIENKDobSBiIqSIcf4O8YKnOlRsX70bxeYSNG6bFJA/KIkoOKxftBmlRWX+DsON0AQcWvVm01cdKjalVT3Mz1Zmw4bFW2A5WYjEFg3R5fKLkNQ6AcYwA6wlnifolWSpWhMiGkMDK9cDwM/vLMBH//4cmiac8xwI4PvXfkXfa3viyW8nIjTc+3KLVPcytx5E9oHjMMVFokOfNlAqTjRFFPRY/AeIooIi2LqGOAv/ms7EoAmEP3MUcq5zZn7pVN6UTtXJIZ+fRFbjRKxISkI/UzYUqXp31iUZiGx4uq0hRKBpm9pNqmcttWLGw5/ij8+XuY0r7NL/IljyCjFyTCwG3ZxXYdy/8x6IAMJuB4rf93IGBTD0hSQZaxVfXRGWZ509IgCcPVeBKHoDkOMghd1U32HRKQs++ROznv4O+Tmnl7q8KLkdHnn/XrTu1sJ/gRFRUCrKL/J3CHXulxkLMeup71BUcHr1noQWDXBRn7auCfw8Sb3rCiz5ZiVsZZ5XGgoJN6LbFR19Fq8vLP9+Nd6f+JnrvXrGzZS1v2/Ca/fMwFOzH/NHaARgx5o9eHf8TOzdlOn6rEHTOIx9+V+46tbL/BgZUf1j8R8gklomQG10GPrFZhgWWSCfVKHFKbClRsF+ZaTXoQC6jSVQsjwnVCEBxu/z8czlffBii3RcasqBJgBrqYS0H2Kx8JtYnMzWIy7RjsG35SHlxjyEhAkIARy2x6GkSMaBXUY0a2uF7PEmqQOSoU/l5xcC00a+hnWLMir02t++ejc0VcO7TzSGrAgMuikfqgPQNAmKTsBuk/DxtBa47937YJR/ArRsD+fXIIUHxlP07at24esXf0L2/j149v/Wo0lrz70VRNF7QOhISDUbk0E+8OObv+HDxz6v8PnutXsx8bL/4t01009PVEVE5ANJrRIgAGgNo+Bo2RBadDgkhwbl6EnoMnMheSl6A4Gik9EjpYvH7T+9PR8fPDqrwue5B09UPZ8AgAE3Xwq71YG0r/+qvIEEjJx4LUIjQqsbcp05eTQPnz31HTKWbsXxQ3ke22mqhhVz1uDw80fRpF2jeoyQAGDX2n/w7yufgWp3fwBz/NBJTL/9bdjK7Bh895V+io6o/nHCPx8510mAtq/fgwmjpkHJtDl7r4tT8+sJQGtpQPHLTSAiK6+8Qz46DsNvBZCq6LVnmd0KIlLGu61XoL09D1Nuao3MneXd0E6Pp2/RoQyv/rgPphgVI1+/EUWv/4NBN+fh328dwu6MUPwxOxYnjuoRHe9Ayk356NynFJISC6nBskon/NuweDOeSH2hWt9Do5ZWDBhWgPAoFccOGLB0bgxKChV8ue0SNIyd6XnHiEcgR4yv1jnqSrG5GP8d9jK2rdzl9nnPARb8++1DiG3oqHQ/KW4uJH1gPcUIdoX5Rbg56V44bJX/O5EVGX2G9sBzcx+v58iIfIsT/vnWuX6fqqpi4CVPwNEywbmqT/lqOpoAVBXGVbsg5xd7P4g/eZnwr6SwFDc2HAO7tXY3MBSdjMH3XIVFs5bCYav8B03HS9vjjeXT/NpdWwiBj6d8iR/emOd1GqIzyYqMu1+4Fbc8PrxOY6OKJvb/L3am74HmYThJeFQYZh/9GMbQwOo5SlQTnPDvPPTVyz9COWhzroRT3m3/1D/lgzaEvJuL0skJkPIcQKgCYToj8TlO/wdNS9DBdk0UHB1DAVVAv74Y+sUWyBYNsAv0iDiOSyKP47mxzZG5KwTua+84/3xgVwhefbgpnv/yAGaO+w0LZBPmvN8AB/e0wZ6McCiKgKoCigIs/DYOvVOK8fSPL8HoYbz9R/8uL9rFWeer6GimEd++k+D2WbhJRXxMxSe0bnHb1rl9YiuzYfmcdPw9fwPsVgdad2uBIfcORHzjmq1TXF0OuwP/HjgNezdmVti2aWUkHruhDd5buAfhkZX0AhAldRITebZs9uoKTwHOpKka1szbgILjZkQ3iKrHyIgomC1cvsNZ+AOnC3/XnxVY+3cAdu6AMcvhnFenzF7VCnn16q5pozzO9D/rqe9qXfgDzmF+6xdt9jrp376MA7CV2lxP/oUQ2LJ8B/78cjkKTljQoHEcUu+5qkarEdTUZ099hx9en1ejfSRZQllxYM31cCE4ui8b21ft9tqm2FyC9F/X44pR/eopKiL/YvEfAE4czcOGuZtcY/TPJmmAfmUR9KuLXG0cF4XAemssHL3CobY1AvMB21WRKH301I8KRQKEgNolFGW3xiH01WMQ0QqGxh7EscMGrFoQBc+FuIS1aSbkHtGjYeMy3PJwGU4c02PBV87CWVWlU/90tl63JAIzJqVj0ifd3Y4i1JPYs/J5ZG7Ndh23NlJuKoUkVf6E9tSZANtqCDUHkpKAw/8cw+ODnkNu1gnIsgxN07Dmtw345qUf8ejH45B6l++7d6386e9KC38A0FQJxw4YsOjbWIy478RZW2VAaenzeMi7E4dPQtHJcHi5ASCEQN6xAhb/ROQz3/26vuKEvmV26A7mQsougEiIhqNdB1g7n7qZXmKFfm82lH3Zfr8JICsyFn62FDdPuR56g971uepQ8fULP1Y5k39VNFVDzkHvQwOsJVas/GktBt05ANZSK5676XWsXbAJik6G6tCg6GTM+/APpN59JR79+H6f9xDIzynA7P/NrfF+ql3lMDI/OH74ZJVtZEXG8cOeh20QBRsONA4AO9P3VDq7rTBKcLQ1QhidKf/MmwPKrjKEP30Uhh/zYL80Ao5uoSidlOD8N6qc+okgSZCP2mFYYoHaPRxypg0NdSXYszEUVRfiEhZ/HwMAsFtlLPkpxuM+QhNY9Pky5OcUnPFZHkTeTVjw6VYoiue7+O27l+DR17Pw7oI9+N+cvbh29AmEhp8uyCRJQr/rmkFCNRK4dhw2qx2PD3oOJ444/0Ouac4vTVM1qA4Nr415H1tW7PB6GPMJC5bNXoXFXy7Hge2Hqj4vgD++WO51uwCw8JvYsz5VAOPVkJS66Y1AnkU3jKrWslZRDdhNmoh8o6TUhsxDJ90Kf/lEIUIWbYKy4zC0JnFwdGgM6E8X1gg1wN6lGUoHd0Rp9wioMdV7ZlMX4zk1VUN2Zi5Wzz3d004IgZfveOecly9UdDKaVmOlFUUnu/L7O+NnYt3CDABwTSRc/s9Fs5bi6+d/9Hosm9WOv+dvwB+fL0PG0m2u3wveLJu9ulrtziTJEiJjI3Dp8N412o/OXXTDqm/ea6qG6IbM9XTh4JP/QHDWkn7CIKHszjjYromC8fs8KPusFcpuSTiTe8jnJ2FPMaHk+cZux5LMKkJfzYZ+Y4lz7oBT+8zo0hADB1c96Q4A7MkIAwBsXxuGshLvxbfm0LDxz60YeHt/AEDh4TcQpjuKw3tbuHoKuBO458ljGDXhOBx2QKcHNA3ollyMWx7KwdRb2uLQXgP63dAbXa9IAoqXVR2wHI+Vc9YgN+vsp+unSQBm/28uul5ecYy9zWrHB49+hoX/t8TtiXCnfh3w+OcTkNQqocI+5U4ereKusZCQf/yMH3RQADkekulJ7/tRnRhwczI++vfnUD2MAZQVGd2u6IS4pJh6joyILhg2BwyrdgKagIiLhNo60fn5mb8JynN6aCQKR7eGrYkDIWstiPnkGJQCzz3iymfx8XVvAVmR8feCjRhw86UAgBU/rsGy2avP7aASkNC8ASZ9Mg6P9n/Ka1NV1RCTGI0TR/Pw55crPC8LKIA5b8zDzVOuR0hYxbHc8z9ejP978hsU5p1efaFhs3g8/P696DOkh8fz52UXQJYkaNWcLktWZEiShKlfPQyDUV/1DuRTzS9qgtbdW2D/loMe/64YQw3oxxszdAHhk/8A0OnSds41YQEIBSh+thFs10cDBgmGBWbPwwEAQAX0aYWATjqd5e0awqcehi7DOZZcEqfnDzi2Q4f5n1XnSbPA5vRwnDimg91evZ8PtrIylOR+jqPr+iBc/z1kWUNElApZrvgf3JQb8zFqgvMmhO5UPpRl5/KCsQkqXvnxCF6c/wSe+n4S5PAh8H6fSgb0vSEpifh7wUZIsud4hQDWLtxU4c69EAIvjHoD8z/+s0JX8F1/78Ej/f6DvOx8j8dt2CwekuTtexKIb1S+IkMoEHYLpLifICmJXvahuhKbGINRU4ZXuk2SJciyhHtevLV+gyKioBYWakCbFg2ciQiAYf1eSJqABMDRsqFz0j8PBASMe3SALKHsEhNyX2oFNcL7Tfm6GCYghIDdakd2UTq+WzoC2e3HY+yWfeh0ewFq1d9AAkY/OwofZryGzv064KK+bSF7yeF6ox79R/TGxsVboFXRe6usqAyb0rZU+HzeB4vw1riP3Qp/wDn7+1PXvYz1f2z2eMwGTeJQ3XmyJUlC7yEX462Vz+OSwRdXax/yvfteuQMAPP5GGz1tFMIi/b96BFF9YfEfAGITY9Dmug4QMmAfEAm1exigSJCKNMiFVXQvkwDl4Kmi8tR/2PQriqAcsFV+00AF8k/oKy3Izz6wtUTBqw83Q4uOZafvHnjRqtWXMKovomHjfNeDi4mvHsYN9x5HZPSZTygEbhqfC83DcGtFEYhtYMElV5ohyzIkOQZSxAQPZ5UByJAi/w0AsNscnp8EnD49DmzLcvto28pdSP91faX7qg4N5hOF+PHN+R4POfjuq6r4QSAha08Idh/4HFLCOsimZyApDbzHSXXqrudvwV3P3QLjWU+FEls0xMt/PIUOvdv6KTIiClY3XtvDmatLbZBzzK7PhSnMfQLAs0iQoMs79ZNNkaDG61F8VXSdxOgtkwlNIKLtCazOeRBl0iFs/DAGv9zWGEf/rl3xpCgyCnLNCA13rjx03yt3QJJljzfx75o2CuFR4R5Xajnbmvkb3d6XlVgxc+rXlbYVwtml8qN/f+4xn18x6lIouqqHId44aRh+sXyB5395grnEz3qkdMXzvz6BuMbuQy/DTKF44I27cONjw/wUGZF/sNt/gDhwtwHqdiNsg6MAVQCKBGGUquy2J2mAfkUh1OYG2IZHA7IEfZoFQvJcrwtI0EdqsJrh8egCQFmKCX/d0BzLj/cAfhHQrS+G8Yd86HZUnLE2NsGGNh22QJLceyyaYlWMfeoYRj2Ui8dvao3MnaGIilXRor3V6/ehOmSkf/cGfvwkDUmtEtC6W3MM+dcjCJVmAuKMZZCUppCiXoJk6A7A2cXLw+rAbmY+8TVeWvAf1/s/Pl8GRadAdVR+R0JTNSz8bAnu/d+/Kt3ed1hPXDywCzKWbvN48+GuF+7ERX2TqxEd1QdJknD7f0fihkeGYN3CDBSbS9C4bSK6Xt6xil4cRES1szYqE6pSAGPWWcv5OdSKEwGeRbJJCP/bgJLuNufvgxDfPb8pX1rYe/81Z++8+GsX4uDyMCx6MBGaQ4LQzhhbKGo22EB1aPj90yXYvX4f4pKi0eyiJnhoxljM/t9cHNuf42oXHhWG0dNGYfhD1wAA2vSo3kS5y79fjfFv3+Pqcv/3bxtQYin1fI1C4MC2QziwLQstuzSvsN0UF4m7nr8VM5/4yuMxul3RCWOm3wadnj+xA0WfIT3wVeYMZCzZhpwDxxEZF4lLBnevdEgIUbDjf5kCQHZpAU4UmhFepkFL0p+esC9EhqNnGHSbSjx2/QcAWAVCZp6ActCK0okJkAtUrw/qJQClmgK5uQIcrHj3XAAoGx0L282xpx8B6CQ4eoXD0TscoW/kwLCk0G2P/3x0EJU9Lyi/GRARpeLFr/fj7n4doDdWPVmOJjTkHTuJHel7sCN9D5Z88xdmTpVxzwtP4KaH4yChEFCaAfoeboVa8rBe+Or5H6o8/rqFGVjy7UpcdetlAJwz+Hoq/MsVniyEEKLSwlBRFDz/6+P48LEvsPD/0tyGDiS2bIjHZo5D9yu7VBkX1b+wyFAMuIk3ZYio7i08uB1hK/dDOWsUme7ISdhjI7zuK0GCca8OuhwFlsGlkOzCd+P6zyj8pdMfnf7zqT90mKDBXiZj0fhEqHbp9AZXI3HWnlWzldqw6+9/AACrTk0m2GdoDzz8/ljkHStAZGwEeg7qCkPI6eWE2/VsDZ1BV2UPgOKCErw97mNM/mw8ACA/xwxJlqrsIZiXXVBp8Q8AN0++DmGRIZj1zGxYTpz+LRRuCsVt/70RIx8d6vNVBujcKYqCnoO6+TsMIr9j8R8A/ik8hrCXsyEftkMqVCFiFFf3P+stsdCdmrTPU0FfnmINiwthG2iClqiHnOWh2/+pHUS4gtIrTAj9vOJEdfb+Ec7CX5Lc87fOuXxg6cQE6DaVQM53FrjR8Q507u19rXpFAeISHZj8dhb2bgtFSZEOoeEOjw859Hpgx/pw13shAKFqmDn1B4RF3Yth425way+EAGxr0bTxbCh6QK3GUsOfPzMbV97SD5IkISQ85PQMSR5EN4zy+kTYGGrEI+/fi3tevBU70vdAUzW0v6Q1YhM5aRwR0YXOpjqAn49Al22vUBorB0/A3rYRYNR77/4vJCiFQMh2PdQoXbVKbLWBCfLJQkheCt7KjnPmZ1p0OOxdW6D7LUuw62sTtLML/zP2kmQBTXPuLykyAAGh1mw+gHULMyArMp6b+3iFbYf/OYb5Hy2GwVh18Q8Af3yxDP96+kYktUxARGxE1UMDAcQ3Pnt1ntMkScKwB1JxzdiB2L5qN4oKitG4bRKX8iOi8wKL/wCw9a+90G11dkMzpFlQdne8a5vaKRQlUxMR9noOhFUACgCt8hsBQgEMv5th7xoK/Zriig1cDQVaRxeiwaGTWAcTzk771ltjoWwvhWGe2RWXo2cYrNdHQ2sTAkgCtsFRCPn2JPoPK8CN46q3eoDDDqiqhBPHDPjmrYYY8+RRQAJUFdiwLBJr/zTBbpfQulMpLrnKgpXzK1+i5cvn5mDI2IGucXdCOCDMjwNl82CUFAy+NRHzv4ircF1nO7o3G5lbs6BpGtb8tt5r4S8rMq4ZMxAAcHDHIRzZm42I6HB0urR9hfF/kTERXmcLJiKiC4+t1IbIn45XXmg7VBhX7oQtuT1ERIhz+RtIld4IkISEkH/0sEUZvT5jFxIQ1jocPe7SYdn6bsCOXOj2HHVrr4YZoLVKgKNxHKDIkC0l0O3PgXw0360HgMlUhIH/XYVwUykOrYw63dW/svNqEmDUwda5GWDUQ796d417J2iqhvRf1yNz60G3J/ALPvkTbz3wMSRJqnLCv3KyJGHlT2sx9L4UzHl1rte2kiyhdbcWaN6xKfJzCrBn/T7IioyL+rZDRHS4W1udXoduV3Sq4ZUREfkXi/8AsP333a4n+4aFFliHRUPE6pxP2gE4LouEpUc4DEssUHaWQb+iEKikh7qkAso/VuhXFFXc6N4SWXtCkPVPCJJTzUhfFIXynw9CAnQbSxA68wSE4jwmAOiXFUK/pBClExNgHxgJtZURgIS8bD1adSyDpjln6z/b3q2hyNwVAtUB5OXosH1dBNYvNUGSBNp2LcZFPUrwn9tbIeufECg6ZyJf6IiFJDf2+OMiP7sAu9buRadL2ztjLnofKPvt1FYV/5qUg6U/R6OkUEFVNwCKzSV476H/g93q/elBfONYXDywCx5OfhI7T3VPBICYhCjc9dwtGHJvitf9iYjowrZ16Q7IJZ4LVrmoDMbFm6EmRkNrYILaOA4INVTe1iZBX9QQ9q6AYcvBCjcBhARAp8DcqhX+3BCGfl0PIF1uBodegX77IQCAGhsBW78OzuR96iaDFmeCrUEU5KzjMGzY73x6D6D4oIT4hnnQ6TWISi5BhOihxpsAISCfsECyOSAXFENtGA17t+bQbz5Y4xsAsk7GXz/+7Sr+t/61E2+O+wgQztUPqn0cRUaJpQTfvPgjDmw/7LWtJEkY/dwovHT721j+/WrXDQa9UY8h9w7Efa/c4Tb8gIjofMPiPwDkO06PGZOKNURMOYySJ5OgtgtxTv4nAQiTobYwwPhtnqsgP5sAIB+zV2difmias4/7mj+iICsCmirBEKqi/9ACpM08FcsZ55FU5/FD38qB2tYIOdfZr37H+nAMa9kVplgHhvzrJEbefxymGBUHdoXg1UeaYu/WMLcIJdn5TyEkvDSuOULCNdjKnHcNVMfpuweV/bg4U0mhs0eCEFag5HOc+dg+tqEDDzx3FK8/2szrMSRJQmlRKfZvOej9ZADunDYK/x32MuxW9/EE+TlmvHn/RygpLMWNkzhjLBERVa6ooGKPPNcw+VMkAEp2AZTsAuj/OQZrv4sgTBVn0hcQUEplqK0TYdMr0O04DKnU5tquxZtg79YCIty576qtLWDU22FtmwQhBHT7c2Dr2w5QZPdJBstvAjSNh5pXBCXrBBxtEuFomYAZz/aBrKiI7ZgNcegEpIIyCJ0CW/cW0JrEQcovgjF9DySbA0IClH050O3LgRYRAq2BCcpxS42+L1mSUFp0eoLhH96YB0WRoTqq98S/nMOuIql1Aj6YOKvK3gJD70vBZ//5Fge2H3Jra7fa8ev7i3Dkn2y88NsTHNNPROctFv8BoNllTXH8ix2u93KuAxETD8HR1gi1cyggAN2WEij7bV6Ocuquv3D+mHD0CIMjOQJCBnSbS6BbXwK5+OykJ0EIAaE6x+TZShX8vTjKuXpeJflRgvPYxu/yoF9VhPIiHgAseTp8/15DLP81GlNnHMTUW1uhtPjs5CidVdRLKKvQpnqatm/k/IN9OyAKK2wfODIfn76UhIITOleMZ5IVGX2G9kD+GUsteZP+0zewW23QPIxb/PQ/3yD17isRGeN9wiYiIrowdTzVW+1MQq+DfNa4dVd3e6sdxlU7UXZ1d2eRXv45BKRTrSRLCSRzKbS4CGiQoMWboCVEQVSYxVzAatcDEFDbNYLaOhGoYsk6e5skOFo0hIg6fRNfUxWc1JIgLk+CYfUu2Ds1hYiOgFRig3HlLuBUwXzmQwipqAxSUcVVgqricKho2qGx6/36RRk1LvwhAeGmMHTu16HSmy9uTWUJm9K24vCeY5VuF5rA+kUZWPd7Bvpe27NmcRARBQjfrRNDtXb/qGFwJBohzvq3ofvHCuPPBTDOLaiy8C+nRckoerspSp5vDNvVJtgHmVA6ORGF37aCbZCpkj3c5/ctMiuVFv6u1hqgX110atiBe1GtaRJyDhnw0oPNUVqsQFN9v1yarMjoOagrEls0PPVJ5d0gFB3w2FuHIMmALLsX7LJORmRMOB544y7nRH/VMPS29Xj95z2IjKl8eIDDpmL59+nVvg4iIrqwNGqVAMB9ehlJ1SqfNw/OAloqs0M5csbEvOJU4S8E9Jv2IyRtK3T7jkE5nAflyEkYMjKh23kYqDCpXflyfKeW4FHkU8vyeSBJQEQIhCkMZ6/hK4QMKICtbzuI2EhAlqDszwY0rcqJA6tNAkLCjLjylktdH2nVmKjvTLIsQZIkPPZ/D8IUF1lle6EJj4W/65iKjIWfLqlRHEREgYTFfwBoGZmA0ieTICIVjz8CKiMkuG4YaKGS8892IOTLk9BtLHbOGaCcmjBIAUonNoSje8Xug+6qDkByeG6lqRKyDxrqqPCXEB4VhgnvjT39oa49gMrH311yZSFe+3EvuiSfngNB0Sm4clQ/vLf2ZSS1SkDPQV2hD9F7PW+4SUWXPsVo370E0z7LRGUzA8o6GScOn6zNZRER0QVAkiQk9GgCABA6BdbebQBV8zpUT0iAnFNwagJAONOPJgBNwNEiAY4WDYBT/QDKj6NknYBux6GqgoHH5XbO5HHlARnQ61w3GZTDJ6s15LA6JNl5RY/NfAChEad/s3To3Qay4v1n65nbO17aHv/74yn0H9EHEdHh6HL5RZCVc/ttoqkacg5Wb5JjIqJAxG7/AUJqa0Txi40QMekwhE1UWYILAI7eYRB6GfqVRZDKhPMpQYkGaUMJ9OtKUHZrLKx3xJ06gQSoAmU3xyIi48g5RFqe3b1F6PvCX9HJuOq2/rjz2ZvPeOoPSLIJInQ4UPoDKuuy0Km3Fa/8HIIC9RMU5RcjLika4VGnZ+yNiA7HyEeGYvYrcz0+BBk1IQeGEHHqeCXo1LsY29e6d+9XHSrCz5oJmIiI6EzRxlAcA2Dv2xZaTGTV2VIAckExlMxcqM3inV31Jec0fCI6DPbuLaE2jYdh1S7XUn4SAN2+bDjaN3IW6HVFCNfNAamKsfQ1+VXQpX9H/OupG3HxVV3cPr/h4SHYtnKXxxPodAo+2fYmhKYhPCqswjK7//rvjXg89fkaRFLJaSQJETFhVTckIgpQfPIfAA4Vn4Ris0L/pwWwV134A85Eah0SBX2686m22/i6Uzk45Ns86DacMcZNkaB2D4MI8X1xXmck4OKBnfFz/ueYMmuCW+HvahL5BKC7CHDNS1xOBuQYSNFvITYhGs06NHYr/Mvd9cItGHrfIACAoggoOg2yIgBJ4KYHcnHz+NN3+R124NLBlUxaJICvX/gBW1bsqLiNiIgueJqm4cDWLKjxJmgNogCdjNLre6P0uktg7dMWamzFOWMkOMfMaw1MFSfnO/X0XouLhKOj+xrzkiYgV2eCPW9d/6siwfXkX4sM9Tr/fnXOojPo8EPu/+H1pdMqFP4A0H9kXwx/6BoA7k/4FZ0MRZEx9etH0KRtEpq2b1yh8AeAHildMWXWBOiN3nv7eSOEwOalOzDn9Xm1PgYRkT+x+D/LjBkz0KJFC4SEhKBPnz5Yu3ZtnZ9z7cm9iJufC+M8c7W6zQkZcLQ3QrejDBBe1viVAcMvBRU/159L8V/1voYQDZLso/5/ArCV2hHqZWy+JEdAivsGUuR/AKU1IIUBciMg/AFIcfMg6Vp4PYWiKHjkg/vw6fanMeqhXFw9Kh+3T8zB52t2YuxTx9x+awkAxpDKn3CUFpbiyWtexLH9ObW4UCIiqi/+yPU5B46jxGqHo9vpdeudw/JkaInRsF3eEY7mDVybyrOoFhPuHHtf2Xq6ACBJcLRsCHFWl/iqnsZDiCq6/gt4L9sl15N/tVXCOfcHdNgcCI30PDRRkiQ8+NbdmDZ3Crpd0QlhkaEwxUUg5V+XY8a6/+HyG5OrPMegOwZg9tGPEdWgsjmQqkcIgY8nf4G0r/+q9TGIiPyF3f7PMHv2bEyaNAkffvgh+vTpg7feegupqanYvXs3Gjas+MTZVzTNhpLPSjzNXedGABAmBaVTEhH6Zo7rKX9lJA3QbS91/yzfAanI2w+Cs1cLrvTIXtvd9EAufvs8HoXmc5/0T9HJaNw2qcp2khQKhN8JKfzOWp+rSYeLMPrxMkAUeI5HAQ7srvxGhKYJ2O0OzH33dzzw5l21joOIiOqOv3K9qmrQEmMBUyXdxmXnBHz2i1tCPm6BVGp11d1aA5PzCbvH8fcAdApEZCikM2a016Kq6J5e5Zh/z9slSUPTNtkoK9AhN7cB1CZxUI/kQT6W77Zn+S8FLdQAudT7xMXRDaNgqOKpvCRJuPS6S3DpdZdUEbtnkTER6H5VZ6z8cU3NVw9wBQJ89fwcXHXbZZCqM3cCEVGA4JP/M7zxxhu49957cffdd6Njx4748MMPERYWhk8//bROz9tUC4GwV11yCwBaCwOK3msGLckAoUhVd6U788eCJmD4rXq9C6qi6Mr/KSDJAoriPOjI+3Nxx79z8Oa8f9C599nL6tT8xKpDw5B7U84x2uqRJB0Qdis8/99CgmpXsOSnit0Jy2kODSt+4Kz/RESByl+5PrFFA6gtGnjuai9JgAAcrRKgJsWc/k1Q3dR56rhCAtTYCGdvAS/tqk9AkjVIkgZZdhbLSc1O4Np/rcSoR5ag48X7IEkCtt5t4ejcDCLk9CS8IjIEtotbQk2K9n4KCRg27uoaxlV71z2QWnXhX0WniMN7juHI3myfxkVEVNf45P8Um82GDRs2YOrUqa7PZFlGSkoK0tMrFnNWqxVWq9X13mKpxtg6D47Nz6n2OP/SCQ0hYp3/2hw9wqDbUurxh4FQAHuvMGeiF4CyswzGH/M9H1/WACFV+btAVgS69yvErRNzseSnaBTm65DQxIbUW/PQrK3zO2nc0oZXf9yHQ3sN2LUpDG8+1hSqo+b3mlp2bY6Oye1qvF9tSeH3QVj/Ahw74D6BoAJA4Ku3e6Kk0OphbydrifftRETkH/7M9Tq9DiI63PsTd1mC2iYR0i4VOOrM10quGY7Ozbwf3GqHZCl1rhik08Heo5XntjV4Uq032HHL+EXYvbkFCk5EwhBiR/tuB9GsTbbrMINv+xsDrs/A0YPxWDavB8xtugNldufDB4MOkCQY/vI+H47eoMPIR4dWO65z1fXyjhj56LX48c3fIMkSRPlkiZIEIQTa9myFfZsPQKviBgHzPRGdb1j8n3LixAmoqoqEhAS3zxMSErBrV8XZZadPn45p06b55NxaNbv7q+2NUC863eXcfrUJId/mQVhFhaf5AgA0wDY8BlKeCsO8Ahjn5kPy0uvu1odz8e3bCZ4buOKVcP2YE+jSpxhd+pz9dP+MGATQtI0Ns99LqFXhL8kS+g7tUa9d6iQ5HIj9CqJ4JlD6DaDlAZAAQz9IEffDUrwRim6JxycGsiKjZdfmlW4jIiL/8meuB1D1U3chAIcG3RlPlGVzCWRLGbRIY+WFuxDO9rIEtVkDONo1gggzej7HyUJIBh1ERIjXGwGSpKHzJfvQIMmMBkmbvYYdEmqFeqgY5pMm55OK0FNP/zUByVwM+USh1/0bt02qdELeunT/a3eidfcWmPPar8jcmgUAaNqhEW6cNAxhplC8MOpNr/vrjToktqy7YSJERHWBxX8tTZ06FZMmTXK9t1gsaNq0qZc9PLvy1n748N+fV7ZSnYsEoPTRhFPdAp2T9IhIBcVPN0L4c0chrMI1+Z9zjJ2AkIDwqYeBMme3fKGVdxI4M9k7P0m5MR+jpzgnqvvmrUQPUTiPPOK+XPQeWOg2V5CmVVw2OP+4DjOfT0Laj7G1+FYAoQlcMvjiWu17LiQ5DFLkwxAREwBhBqQQ55wCAIaNi8eCT/70uK+marjuwcH1FSoREdUhX+Z6AJDL7NAU2fPkfQB0uw5DcpzxVCA8FMbjZSgz6iGMyumJ+s5Iwo52SXB0bOJM07J0+iZDeVIuf19cBuPKnUB8JKyXdkDl8/c42zZsnId+gysW/WfPE+iwStj2VRRWrOgDJLg3kkqsMKzZ47V3o6zI6Hd9by8t6oYkSRh0xwAMumMAis3FEAIIjwqDJEmw2+yIamCC5WShq1eAW8w6GSn/uhzhnoZWEBEFKBb/p8THx0NRFOTkuM/UnpOTg8TEisWw0WiE0ejlznoNxCbEIDwqHMX5lT9FFxLguCQMWjPn+aR8Ffpf86FF6SAa61H0dBIMywoRtdkCk96GTpcUI6GJDX/Nj0Z2lhFh0SoGjszHoFF5SPshFgu+ikVZiQIAaNpWwtinTqDvIAug64Y7pw1CVOyX+ObteJhPnv7rIUkCbbuW4MYHjqPH5YVwOIDSIhmh4QKKTuBophHzZsUhc1cIYhs6YD6pQ8aqCNzwyDA89X07zPvwD2z7aycc9tM/aM7sanc2RSejZZfm6HxZB598x7UhSTIguY/vb3NxS9zx9E348rk5lXYVvOq2y9B/ZB9/hEtERFXwZ64HgMFDe2DBsh2Vz7SvCcDugO6Ac3lZcepJvmgYC8UhoD9wEqqsQU00AToFkqUU8slCaA1MELGRgKpBzjVD2ZsNRIXB3iYROLVSjqwJRJ+wANuyENOyIVLvvgInO/6GtMVNcCI7+owgBMJNpeh1+U60bbsPiuSArUiCJAG6UAGrWcauH004sCQM4QkqVKuEI3+HIqlJc7zzwwT8smgzVi/djuITFuiO5kPOOg4ZnqctkGQJOoMOQ+6rn7l9PDm714HeoMeT30zEf4a+BE3VoJ2xcoKsyEhs0RBjpt9e32ESEZ0zSYhzWeQ1uPTp0we9e/fGu+++C8C5Jm+zZs0wYcIEPPHEE173tVgsiIqKgtlshslU8yVkcg8dx+0tHgTE6fvwZ/5Ti1Ocxb8mYL02Cmq/SACnnvBDQt/IY3ilZTr0kgZJqvxJ/JH9emxZ3wNX3X4VLCdDERJzJaIaVOzmL6wrYc8dj50bFJSVyGjSxorEplYU5iv4z+2tsGez8063rMjQVA3dr+qMjCXbICsyyv86CU3gmrED8cj790LROW80lBaVIu3rlfjrx3SUFJahZZdmsJXakPb1X1B0MlSH5iqoG7dNwqtpz6BBk7gaf5f1Yfn3qzH7lV/wz8b9AIBGbRIxcuK1uHbcIMhenugQEdWnc81NwcifuV7TNFx78xuwqFrF4t/mgFRQ5LzxLADppAX63UddvwPKSeX/IwDJBCgNJTj2nm4hJCAsIgS3PHEDdKYwdB3QEe06NoV81moBJY5s/HVkHA5kmVFkjkBYZAkaJOUDEFj5XANs/ybq9AkF0OnSdtizYT8cNhWQTs3irwm07t4CLy14ErGJMa5rXLcwA4tmLUVu1gnEN4pF0w6N8cPrv0LThKuQlmQJBqMe0+ZOQc9B3Wr8XdaHfzbuxzcv/YTVc9dC0wTCo8IwZOxA3DL1BphiI/0dHhERgJrlJhb/Z5g9ezZGjx6Njz76CL1798Zbb72F77//Hrt27aowPvBsvviBlZ9TgPF9n0TuoeOQ9YA+AghpK0GfBKAIkI7akZepwHqFCbaroiCidIixheDWDi1wy8UHIKkHIaQQKCFDsHZJY/zy3g8I0W1ESLiK7KPN0P/Gm3D9+MGuYtwboZmB0p8gbJsASYIq98bcT/T48a00nDySBwC4eGAX3PLEDegxsAsO7T6CxV8sx4mjeYhNiMagOwegecfqdY3cvX4fFny8GAd3HkZEVDgGjLoUA25KhuGMGYMDVbGlBJqqISI6nMv9EFHAYfFfkb9zvaZpeHza91izcjcgBCQ4oFiLAaMEzaGHLcSAotaRMByzIHxrLnTmMhgiQtAtpQvaxEVj35aD0GQVF1/TGd37dMFPb87H8jnpEKqAIVSPofcOwq1PjkBMw6iqYxF2HCtejqPFS6GKMpgMbZC7sgV+eHEFdq/bC8A5Hn/kxKEYev8gFBUUY/Hny5G5NQuGUAMuvf4S9EjpUq2b3ieOnMSCT9Kwedl2QAJ6pHTFkLEDEZMQXavvsT7ZymwoK7YiPDoMilL1bygiovrE4v8cvPfee3j11VeRnZ2N7t2745133kGfPlV34/blDywhBPKPm6HTyYiMiYQqNBQ7rAhXFCgyIEnV74JYWlwG1a66xrGdKyEEis0l0Bt1MIb6riskERH5Hov/ygVCrgeAoqJSWIvLEB1vgiRJrvwqSRIMIYZq5227zY7SojKER/muOC0tKoWmaggz+eb3AxER1Q0W/37AH1hERBRomJt8i98nEREFmprkJg5OJiIiIiIiIgpyLP6JiIiIiIiIghyLfyIiIiIiIqIgx+KfiIiIiIiIKMix+CciIiIiIiIKciz+iYiIiIiIiIIci38iIiIiIiKiIMfin4iIiIiIiCjIsfgnIiIiIiIiCnIs/omIiIiIiIiCnM7fAQQLIQQAwGKx+DkSIiIip/KcVJ6j6Nww1xMRUaCpSa5n8e8jhYWFAICmTZv6ORIiIiJ3hYWFiIqK8ncY5z3meiIiClTVyfWS4OMAn9A0DUePHkVkZCQkSTqnY1ksFjRt2hSHDh2CyWTyUYTnhwv52oEL+/p57bx2XrvvCSFQWFiIRo0aQZY50u9cMdf7zoV8/bx2Xjuv/cIRaLmeT/59RJZlNGnSxKfHNJlMF9z/QcpdyNcOXNjXz2vntV9o6vra+cTfd5jrfe9Cvn5eO6/9QsNr93+u52MAIiIiIiIioiDH4p+IiIiIiIgoyLH4D0BGoxHPPPMMjEajv0OpdxfytQMX9vXz2nntF5oL+dqJ//4v5OvntfPaLzS89sC5dk74R0RERERERBTk+OSfiIiIiIiIKMix+CciIiIiIiIKciz+iYiIiIiIiIIci38iIiIiIiKiIMfiPwDNmDEDLVq0QEhICPr06YO1a9f6OySvVqxYgWHDhqFRo0aQJAlz58512y6EwNNPP42kpCSEhoYiJSUF//zzj1ubvLw83H777TCZTIiOjsaYMWNQVFTk1mbLli3o378//r+9ew+Kqn7/AP5eWHYB5aZcFhVQQ0VRCEFpU9QGJrzMdHPKcaw0u4hC6ugYlhbVmDh2mSknUbuok46kmWGJGuOFghQDua0Q3jDNEbGMm3fc5/eHP86XE+j3G7fFw/s1c2bc83nY8zzL4fOcjwt7HB0d4efnh5UrVzbJZdu2bQgKCoKjoyOGDRuG9PT0Nq+3seTkZIwYMQIuLi7w9vbGE088gbKyMlXM9evXER8fj549e6J79+6YPHkyLl68qIo5e/YsJk2aBGdnZ3h7e2PRokWor69XxRw8eBDDhw+H0WhEYGAgNmzY0CSfjjx3UlJSEBISAldXV7i6usJsNmP37t3KuFbrbs6KFSug0+kwf/58ZZ9W63/77beh0+lUW1BQkDKu1bobnD9/Hs8++yx69uwJJycnDBs2DLm5ucq4luc7alu2nrf+LfZ69nr2evb6BlqtuzHN9nuhTiU1NVUMBoN8+eWXcuzYMXn55ZfF3d1dLl68aOvU7io9PV2WLFki3377rQCQHTt2qMZXrFghbm5u8t1330lhYaE89thj0q9fP7l27ZoSM378eAkNDZXDhw/Lzz//LIGBgTJ16lRlvLq6Wnx8fGTatGlisVhky5Yt4uTkJGvXrlVisrOzxd7eXlauXCklJSWydOlScXBwkOLi4narPTY2VtavXy8Wi0UKCgpk4sSJ4u/vL3V1dUpMXFyc+Pn5yb59+yQ3N1ceeughefjhh5Xx+vp6GTp0qMTExEh+fr6kp6eLp6envP7660rM6dOnxdnZWRYsWCAlJSWyatUqsbe3lz179igxHX3u7Ny5U3bt2iXHjx+XsrIyeeONN8TBwUEsFoum6/6nI0eOSN++fSUkJETmzZun7Ndq/UlJSRIcHCwXLlxQtkuXLmm+bhGRy5cvS0BAgMyYMUNycnLk9OnTsnfvXjl58qQSo+X5jtqOreetlmCvZ69nr2ev13rdDbTc77n472RGjhwp8fHxyuPbt29Lr169JDk52YZZ/e/+eUFgtVrFZDLJ+++/r+yrqqoSo9EoW7ZsERGRkpISASC//vqrErN7927R6XRy/vx5ERFZvXq1eHh4yI0bN5SYxMREGTRokPL4mWeekUmTJqnyiYyMlFmzZrVpjfdSWVkpACQzM1NE7tTq4OAg27ZtU2JKS0sFgBw6dEhE7lxQ2dnZSUVFhRKTkpIirq6uSr2vvfaaBAcHq441ZcoUiY2NVR53hnPHw8NDPv/88y5Td21trQwYMEAyMjJk7NixygWBlutPSkqS0NDQZse0XLfInTln9OjRdx3vavMdtVxnmK9bg72evZ69Xtvf967c60W03e/5a/+dyM2bN5GXl4eYmBhln52dHWJiYnDo0CEbZtZy5eXlqKioUNXk5uaGyMhIpaZDhw7B3d0dERERSkxMTAzs7OyQk5OjxIwZMwYGg0GJiY2NRVlZGf7++28lpvFxGmI68rWrrq4GAPTo0QMAkJeXh1u3bqnyCgoKgr+/v6r+YcOGwcfHR5V3TU0Njh07psTcqzZbnzu3b99Gamoqrly5ArPZ3GXqjo+Px6RJk5rkqPX6T5w4gV69eqF///6YNm0azp49C0D7de/cuRMRERF4+umn4e3tjbCwMHz22WfKeFeb76hlbD1vtYeudu6z17PXA9r/vnfVXg9ou99z8d+J/Pnnn7h9+7bqBwUAfHx8UFFRYaOsWqch73vVVFFRAW9vb9W4Xq9Hjx49VDHNPUfjY9wtpqNeO6vVivnz52PUqFEYOnSokpPBYIC7u/td82pNbTU1Nbh27ZrNzp3i4mJ0794dRqMRcXFx2LFjB4YMGaL5ugEgNTUVR48eRXJycpMxLdcfGRmJDRs2YM+ePUhJSUF5eTmioqJQW1ur6boB4PTp00hJScGAAQOwd+9ezJ49G3PnzsXGjRtV+XeF+Y5ajr3+P+7Hc5+9nr2+gZbr78q9HtB2v9e36KuIqIn4+HhYLBZkZWXZOpUOM2jQIBQUFKC6uhrffPMNpk+fjszMTFun1e7OnTuHefPmISMjA46OjrZOp0NNmDBB+XdISAgiIyMREBCArVu3wsnJyYaZtT+r1YqIiAgsX74cABAWFgaLxYI1a9Zg+vTpNs6OiDoCez17fVfQlXs9oO1+z3f+OxFPT0/Y29s3+bTMixcvwmQy2Sir1mnI+141mUwmVFZWqsbr6+tx+fJlVUxzz9H4GHeL6YjXLiEhAT/88AMOHDiAPn36KPtNJhNu3ryJqqqqu+bVmtpcXV3h5ORks3PHYDAgMDAQ4eHhSE5ORmhoKD7++GPN152Xl4fKykoMHz4cer0eer0emZmZ+OSTT6DX6+Hj46Pp+htzd3fHwIEDcfLkSc1/3319fTFkyBDVvsGDByu/CtlV5jtqnc7wc9vWusq5z17PXs9er/1eD2i733Px34kYDAaEh4dj3759yj6r1Yp9+/bBbDbbMLOW69evH0wmk6qmmpoa5OTkKDWZzWZUVVUhLy9Pidm/fz+sVisiIyOVmJ9++gm3bt1SYjIyMjBo0CB4eHgoMY2P0xDTnq+diCAhIQE7duzA/v370a9fP9V4eHg4HBwcVHmVlZXh7NmzqvqLi4tVE0RGRgZcXV2Viee/1dZZzh2r1YobN25ovu7o6GgUFxejoKBA2SIiIjBt2jTl31quv7G6ujqcOnUKvr6+mv++jxo1qsntvY4fP46AgAAA2p/vqG10hp/btqb1c5+9Xo29nr1e63Vrut+36GMCqd2kpqaK0WiUDRs2SElJibzyyivi7u6u+rTMzqa2tlby8/MlPz9fAMhHH30k+fn58vvvv4vInVthuLu7S1pamhQVFcnjjz/e7K0wwsLCJCcnR7KysmTAgAGqW2FUVVWJj4+PPPfcc2KxWCQ1NVWcnZ2b3ApDr9fLBx98IKWlpZKUlNTut/+ZPXu2uLm5ycGDB1W3Q7l69aoSExcXJ/7+/rJ//37Jzc0Vs9ksZrNZGW+4Hcqjjz4qBQUFsmfPHvHy8mr2diiLFi2S0tJS+fTTT5u9HUpHnjuLFy+WzMxMKS8vl6KiIlm8eLHodDr58ccfNV333TT+BGAR7da/cOFCOXjwoJSXl0t2drbExMSIp6enVFZWarpukTu3etLr9fLee+/JiRMnZPPmzeLs7CybNm1SYrQ831Hb6Szz1r/BXs9ez17PXt8Ver2Itvs9F/+d0KpVq8Tf318MBoOMHDlSDh8+bOuU7unAgQMCoMk2ffp0EblzO4w333xTfHx8xGg0SnR0tJSVlame46+//pKpU6dK9+7dxdXVVV544QWpra1VxRQWFsro0aPFaDRK7969ZcWKFU1y2bp1qwwcOFAMBoMEBwfLrl272q1uEWm2bgCyfv16JebatWsyZ84c8fDwEGdnZ3nyySflwoULquc5c+aMTJgwQZycnMTT01MWLlwot27dUsUcOHBAHnzwQTEYDNK/f3/VMRp05Lkzc+ZMCQgIEIPBIF5eXhIdHa1cDIhot+67+ecFgVbrnzJlivj6+orBYJDevXvLlClTVPe91WrdDb7//nsZOnSoGI1GCQoKknXr1qnGtTzfUdvqDPPWv8Fez17PXs9e30CrdTem1X6vExFp2e8MEBEREREREdH9gH/zT0RERERERKRxXPwTERERERERaRwX/0REREREREQax8U/ERERERERkcZx8U9ERERERESkcVz8ExEREREREWkcF/9EREREREREGsfFPxEREREREZHGcfFPRK0yY8YM6HS6JtvJkydtnRoRERG1AfZ6Im3Q2zoBIrr/jR8/HuvXr1ft8/LyUj2+efMmDAZDR6ZFREREbYS9nuj+x3f+iajVjEYjTCaTaouOjkZCQgLmz58PT09PxMbGAgAyMzMxcuRIGI1G+Pr6YvHixaivrwcAnDlzptl3FsaNG6ccKysrC1FRUXBycoKfnx/mzp2LK1euKON9+/bF8uXLMXPmTLi4uMDf3x/r1q3r0NeDiIhIa9jrie5/XPwTUbvZuHEjDAYDsrOzsWbNGpw/fx4TJ07EiBEjUFhYiJSUFHzxxRdYtmwZAMDPzw8XLlxQtvz8fPTs2RNjxowBAJw6dQrjx4/H5MmTUVRUhK+//hpZWVlISEhQHffDDz9EREQE8vPzMWfOHMyePRtlZWUdXj8REZHWsdcT3T90IiK2ToKI7l8zZszApk2b4OjoqOybMGECLl26hJqaGhw9elTZv2TJEmzfvh2lpaXQ6XQAgNWrVyMxMRHV1dWws/vP/0dev34d48aNg5eXF9LS0mBnZ4eXXnoJ9vb2WLt2rRKXlZWFsWPH4sqVK3B0dETfvn0RFRWFr776CgAgIjCZTHjnnXcQFxfX3i8HERGR5rDXE2kD/+afiFrtkUceQUpKivK4W7dumDp1KsLDw1VxpaWlMJvNysUAAIwaNQp1dXX4448/4O/vr+yfOXMmamtrkZGRoVwoFBYWoqioCJs3b1biRARWqxXl5eUYPHgwACAkJEQZ1+l0MJlMqKysbNuiiYiIuhD2eqL7Hxf/RNRq3bp1Q2BgYLP7W2LZsmXYu3cvjhw5AhcXF2V/XV0dZs2ahblz5zb5msYXEw4ODqoxnU4Hq9XaolyIiIiIvZ5IC7j4J6IOM3jwYGzfvh0iorwjkJ2dDRcXF/Tp0wcAsH37drz77rvYvXs3HnjgAdXXDx8+HCUlJc1efBAREZHtsdcTdV78wD8i6jBz5szBuXPn8Oqrr+K3335DWloakpKSsGDBAtjZ2cFiseD5559HYmIigoODUVFRgYqKCly+fBkAkJiYiF9++QUJCQkoKCjAiRMnkJaW1uRDgIiIiMg22OuJOi8u/omow/Tu3Rvp6ek4cuQIQkNDERcXhxdffBFLly4FAOTm5uLq1atYtmwZfH19le2pp54CcOfv+zIzM3H8+HFERUUhLCwMb731Fnr16mXLsoiIiOj/sdcTdV78tH8iIiIiIiIijeM7/0REREREREQax8U/ERERERERkcZx8U9ERERERESkcVz8ExEREREREWkcF/9EREREREREGsfFPxEREREREZHGcfFPREREREREpHFc/BMRERERERFpHBf/RERERERERBrHxT8RERERERGRxnHxT0RERERERKRxXPwTERERERERadz/AQYzlyKE7lJzAAAAAElFTkSuQmCC\n" + }, + "metadata": {} + } + ], + "source": [ + "# Your code here:\n", + "fig, axs = plt.subplots(1, 2, figsize=(12, 5))\n", + "\n", + "axs[0].scatter(customers['Frozen'], customers['Delicassen'], c=customers['labels'], cmap='viridis', label='labels')\n", + "\n", + "axs[0].set_xlabel('Frozen')\n", + "axs[0].set_ylabel('Delicassen')\n", + "\n", + "axs[1].scatter(customers['Frozen'], customers['Delicassen'], c=customers['label_dbscan'], cmap='viridis', label='label_dbscan')\n", + "\n", + "axs[1].set_xlabel('Frozen')\n", + "axs[1].set_ylabel('Delicassen')\n", + "\n", + "\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "2OwZX0fr2ael" + }, + "source": [ + "Let's use a groupby to see how the mean differs between the groups. Group `customers` by `labels` and `labels_DBSCAN` respectively and compute the means for all columns." + ] + }, + { + "cell_type": "code", + "execution_count": 20, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 206 + }, + "id": "hAVssEbg2ael", + "outputId": "c014e754-5410-43c3-920f-3ca122c23370" + }, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery \\\n", + "labels \n", + "0 1.004762 3.000000 13423.571429 3172.000000 3746.371429 \n", + "1 1.846154 2.615385 25770.769231 35160.384615 41977.384615 \n", + "2 2.000000 2.672000 7877.640000 8913.512000 14212.624000 \n", + "3 1.054348 1.315217 12407.130435 3401.771739 4234.130435 \n", + "\n", + " Frozen Detergents_Paper Delicassen label_dbscan \n", + "labels \n", + "0 3427.100000 768.219048 1267.142857 0.309524 \n", + "1 6844.538462 19867.384615 7880.307692 -1.000000 \n", + "2 1339.280000 6149.592000 1537.168000 -0.520000 \n", + "3 4082.282609 864.739130 1198.402174 1.326087 " + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassenlabel_dbscan
labels
01.0047623.00000013423.5714293172.0000003746.3714293427.100000768.2190481267.1428570.309524
11.8461542.61538525770.76923135160.38461541977.3846156844.53846219867.3846157880.307692-1.000000
22.0000002.6720007877.6400008913.51200014212.6240001339.2800006149.5920001537.168000-0.520000
31.0543481.31521712407.1304353401.7717394234.1304354082.282609864.7391301198.4021741.326087
\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 20 + } + ], + "source": [ + "# Your code here:\n", + "mean_kmeans = customers.groupby('labels').mean()\n", + "mean_kmeans_df = pd.DataFrame(mean_kmeans)\n", + "\n", + "mean_dbscan = customers.groupby('label_dbscan').mean()\n", + "mean_dbscan_df = pd.DataFrame(mean_dbscan)\n", + "\n", + "mean_kmeans_df\n" + ] + }, + { + "cell_type": "code", + "source": [ + "mean_dbscan_df" + ], + "metadata": { + "id": "1H_p-JeC4A2u", + "outputId": "01d49c1e-f5f8-4fbd-fadb-3dc59b5b7126", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 362 + } + }, + "execution_count": 21, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " Channel Region Fresh Milk Grocery \\\n", + "label_dbscan \n", + "-1 1.490196 2.45098 13729.564706 8359.886275 11484.364706 \n", + " 0 2.000000 3.00000 11401.000000 3536.200000 6729.000000 \n", + " 1 1.000000 3.00000 9128.823077 1862.438462 2423.530769 \n", + " 2 1.000000 3.00000 41446.600000 1421.200000 2167.600000 \n", + " 3 2.000000 3.00000 2357.200000 10224.800000 11862.800000 \n", + " 4 2.000000 3.00000 3068.142857 6095.857143 10355.000000 \n", + " 5 1.000000 1.00000 9837.863636 1751.863636 2161.681818 \n", + " 6 1.000000 1.00000 6823.000000 1943.000000 2372.666667 \n", + " 7 1.000000 2.00000 7494.600000 1537.600000 3383.000000 \n", + "\n", + " Frozen Detergents_Paper Delicassen labels \n", + "label_dbscan \n", + "-1 3795.003922 4356.800000 2080.470588 1.592157 \n", + " 0 720.200000 3561.000000 258.200000 2.000000 \n", + " 1 2131.430769 496.576923 741.946154 0.000000 \n", + " 2 1276.400000 416.200000 784.800000 0.000000 \n", + " 3 609.600000 5780.000000 572.800000 2.000000 \n", + " 4 994.000000 4675.571429 1826.428571 2.000000 \n", + " 5 1513.181818 333.954545 669.045455 3.000000 \n", + " 6 6780.333333 386.500000 613.666667 3.000000 \n", + " 7 2575.400000 227.400000 941.000000 3.000000 " + ], + "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", + "
ChannelRegionFreshMilkGroceryFrozenDetergents_PaperDelicassenlabels
label_dbscan
-11.4901962.4509813729.5647068359.88627511484.3647063795.0039224356.8000002080.4705881.592157
02.0000003.0000011401.0000003536.2000006729.000000720.2000003561.000000258.2000002.000000
11.0000003.000009128.8230771862.4384622423.5307692131.430769496.576923741.9461540.000000
21.0000003.0000041446.6000001421.2000002167.6000001276.400000416.200000784.8000000.000000
32.0000003.000002357.20000010224.80000011862.800000609.6000005780.000000572.8000002.000000
42.0000003.000003068.1428576095.85714310355.000000994.0000004675.5714291826.4285712.000000
51.0000001.000009837.8636361751.8636362161.6818181513.181818333.954545669.0454553.000000
61.0000001.000006823.0000001943.0000002372.6666676780.333333386.500000613.6666673.000000
71.0000002.000007494.6000001537.6000003383.0000002575.400000227.400000941.0000003.000000
\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ] + }, + "metadata": {}, + "execution_count": 21 + } + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "cnkR45Ka2aem" + }, + "source": [ + "Which algorithm appears to perform better?" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "FzCNHxSK2aem" + }, + "outputs": [], + "source": [ + "# Your observations here" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "BmuHBhXI2aem" + }, + "source": [ + "# Bonus Challenge 1 - Changing K-Means Number of Clusters\n", + "\n", + "As we mentioned earlier, we don't need to worry about the number of clusters with DBSCAN because it automatically decides that based on the parameters we send to it. But with K-Means, we have to supply the `n_clusters` param (if you don't supply `n_clusters`, the algorithm will use `8` by default). You need to know that the optimal number of clusters differs case by case based on the dataset. K-Means can perform badly if the wrong number of clusters is used.\n", + "\n", + "In advanced machine learning, data scientists try different numbers of clusters and evaluate the results with statistical measures (read [here](https://en.wikipedia.org/wiki/Cluster_analysis#External_evaluation)). We are not using statistical measures today but we'll use our eyes instead. In the cells below, experiment with different number of clusters and visualize with scatter plots. What number of clusters seems to work best for K-Means?" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "9XmCI7kk2aem" + }, + "outputs": [], + "source": [ + "# Your code here" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "xFPvbQGy2aen" + }, + "outputs": [], + "source": [ + "# Your comment here" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "c56s2VFN2aen" + }, + "source": [ + "# Bonus Challenge 2 - Changing DBSCAN `eps` and `min_samples`\n", + "\n", + "Experiment changing the `eps` and `min_samples` params for DBSCAN. See how the results differ with scatter plot visualization." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "xl8rky2L2aen" + }, + "outputs": [], + "source": [ + "# Your code here" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "DZY-t9I32aen" + }, + "outputs": [], + "source": [ + "# Your comment here" + ] + } + ], + "metadata": { + "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": "3.6.8" + }, + "colab": { + "provenance": [] + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} \ No newline at end of file diff --git a/your-code/main.ipynb b/your-code/main.ipynb deleted file mode 100755 index 9a2d1ae..0000000 --- a/your-code/main.ipynb +++ /dev/null @@ -1,386 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Before your start:\n", - "- Read the README.md file\n", - "- Comment as much as you can and use the resources in the README.md file\n", - "- Happy learning!" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Import your libraries:\n", - "\n", - "%matplotlib inline\n", - "\n", - "import matplotlib.pyplot as plt\n", - "import numpy as np\n", - "import pandas as pd" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 1 - Import and Describe the Dataset\n", - "\n", - "In this lab, we will use a dataset containing information about customer preferences. We will look at how much each customer spends in a year on each subcategory in the grocery store and try to find similarities using clustering.\n", - "\n", - "The origin of the dataset is [here](https://archive.ics.uci.edu/ml/datasets/wholesale+customers)." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# loading the data:\n", - "customers = pd.read_csv('../Wholesale customers data.csv')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Explore the dataset with mathematical and visualization techniques. What do you find?\n", - "\n", - "Checklist:\n", - "\n", - "* What does each column mean?\n", - "* Any categorical data to convert?\n", - "* Any missing data to remove?\n", - "* Column collinearity - any high correlations?\n", - "* Descriptive statistics - any outliers to remove?\n", - "* Column-wise data distribution - is the distribution skewed?\n", - "* Etc.\n", - "\n", - "Additional info: Over a century ago, an Italian economist named Vilfredo Pareto discovered that roughly 20% of the customers account for 80% of the typical retail sales. This is called the [Pareto principle](https://en.wikipedia.org/wiki/Pareto_principle). Check if this dataset displays this characteristic." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your observations here" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 2 - Data Cleaning and Transformation\n", - "\n", - "If your conclusion from the previous challenge is the data need cleaning/transformation, do it in the cells below. However, if your conclusion is the data need not be cleaned or transformed, feel free to skip this challenge. But if you do choose the latter, please provide rationale." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your comment here" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 3 - Data Preprocessing\n", - "\n", - "One problem with the dataset is the value ranges are remarkably different across various categories (e.g. `Fresh` and `Grocery` compared to `Detergents_Paper` and `Delicassen`). If you made this observation in the first challenge, you've done a great job! This means you not only completed the bonus questions in the previous Supervised Learning lab but also researched deep into [*feature scaling*](https://en.wikipedia.org/wiki/Feature_scaling). Keep on the good work!\n", - "\n", - "Diverse value ranges in different features could cause issues in our clustering. The way to reduce the problem is through feature scaling. We'll use this technique again with this dataset.\n", - "\n", - "#### We will use the `StandardScaler` from `sklearn.preprocessing` and scale our data. Read more about `StandardScaler` [here](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html#sklearn.preprocessing.StandardScaler).\n", - "\n", - "*After scaling your data, assign the transformed data to a new variable `customers_scale`.*" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your import here:\n", - "\n", - "from sklearn.preprocessing import StandardScaler\n", - "\n", - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 4 - Data Clustering with K-Means\n", - "\n", - "Now let's cluster the data with K-Means first. Initiate the K-Means model, then fit your scaled data. In the data returned from the `.fit` method, there is an attribute called `labels_` which is the cluster number assigned to each data record. What you can do is to assign these labels back to `customers` in a new column called `customers['labels']`. Then you'll see the cluster results of the original data." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "scrolled": true - }, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Count the values in `labels`." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 5 - Data Clustering with DBSCAN\n", - "\n", - "Now let's cluster the data using DBSCAN. Use `DBSCAN(eps=0.5)` to initiate the model, then fit your scaled data. In the data returned from the `.fit` method, assign the `labels_` back to `customers['labels_DBSCAN']`. Now your original data have two labels, one from K-Means and the other from DBSCAN." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "scrolled": true - }, - "outputs": [], - "source": [ - "# Your code here\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Count the values in `labels_DBSCAN`." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Challenge 6 - Compare K-Means with DBSCAN\n", - "\n", - "Now we want to visually compare how K-Means and DBSCAN have clustered our data. We will create scatter plots for several columns. For each of the following column pairs, plot a scatter plot using `labels` and another using `labels_DBSCAN`. Put them side by side to compare. Which clustering algorithm makes better sense?\n", - "\n", - "Columns to visualize:\n", - "\n", - "* `Detergents_Paper` as X and `Milk` as y\n", - "* `Grocery` as X and `Fresh` as y\n", - "* `Frozen` as X and `Delicassen` as y" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Visualize `Detergents_Paper` as X and `Milk` as y by `labels` and `labels_DBSCAN` respectively" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Visualize `Grocery` as X and `Fresh` as y by `labels` and `labels_DBSCAN` respectively" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Visualize `Frozen` as X and `Delicassen` as y by `labels` and `labels_DBSCAN` respectively" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Let's use a groupby to see how the mean differs between the groups. Group `customers` by `labels` and `labels_DBSCAN` respectively and compute the means for all columns." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here:\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Which algorithm appears to perform better?" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your observations here" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Bonus Challenge 1 - Changing K-Means Number of Clusters\n", - "\n", - "As we mentioned earlier, we don't need to worry about the number of clusters with DBSCAN because it automatically decides that based on the parameters we send to it. But with K-Means, we have to supply the `n_clusters` param (if you don't supply `n_clusters`, the algorithm will use `8` by default). You need to know that the optimal number of clusters differs case by case based on the dataset. K-Means can perform badly if the wrong number of clusters is used.\n", - "\n", - "In advanced machine learning, data scientists try different numbers of clusters and evaluate the results with statistical measures (read [here](https://en.wikipedia.org/wiki/Cluster_analysis#External_evaluation)). We are not using statistical measures today but we'll use our eyes instead. In the cells below, experiment with different number of clusters and visualize with scatter plots. What number of clusters seems to work best for K-Means?" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your comment here" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Bonus Challenge 2 - Changing DBSCAN `eps` and `min_samples`\n", - "\n", - "Experiment changing the `eps` and `min_samples` params for DBSCAN. See how the results differ with scatter plot visualization." - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your code here" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": {}, - "outputs": [], - "source": [ - "# Your comment here" - ] - } - ], - "metadata": { - "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": "3.6.8" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -}