{ "cells": [ { "cell_type": "code", "source": [ "import pandas as pd\n", "\n", "pd.set_option('display.max_rows', None)\n", "pd.set_option('display.max_columns', None)\n", "\n", "# Read the CSV file into a DataFrame\n", "df = pd.read_csv('Mall_Customers.csv')\n", "\n", "# Display the first 5 rows\n", "print(df.head().to_markdown(index=False, numalign=\"left\", stralign=\"left\"))\n", "\n", "# Print the column names and their data types\n", "print(df.info())" ], "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "| CustomerID | Gender | Age | Annual Income (k$) | Spending Score (1-100) |\n", "|:-------------|:---------|:------|:---------------------|:-------------------------|\n", "| 1 | Male | 19 | 15 | 39 |\n", "| 2 | Male | 21 | 15 | 81 |\n", "| 3 | Female | 20 | 16 | 6 |\n", "| 4 | Female | 23 | 16 | 77 |\n", "| 5 | Female | 31 | 17 | 40 |\n", "\n", "RangeIndex: 200 entries, 0 to 199\n", "Data columns (total 5 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 CustomerID 200 non-null int64 \n", " 1 Gender 200 non-null object\n", " 2 Age 200 non-null int64 \n", " 3 Annual Income (k$) 200 non-null int64 \n", " 4 Spending Score (1-100) 200 non-null int64 \n", "dtypes: int64(4), object(1)\n", "memory usage: 7.9+ KB\n", "None\n" ] } ], "execution_count": null, "metadata": { "id": "v5BJ2oxKXGC4", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "35e2ff0e-473d-4e42-d86b-51f5eb890ef1" } }, { "cell_type": "markdown", "source": [ "\n", "I'll use the `Annual Income (k$)` and `Spending Score (1-100)` columns to demonstrate how DBSCAN works. But first, I'll plot these two columns to get an overview of the data." ], "metadata": { "id": "UddW_d01XGC5" } }, { "cell_type": "code", "source": [ "import altair as alt\n", "\n", "# Extract the features\n", "X = df[['Annual Income (k$)', 'Spending Score (1-100)']]\n", "\n", "# Create a scatter plot\n", "chart = alt.Chart(X).mark_point().encode(\n", " x='Annual Income (k$)',\n", " y='Spending Score (1-100)',\n", " tooltip=['Annual Income (k$)', 'Spending Score (1-100)']\n", ").properties(title='Scatter Plot of Spending Score vs. Annual Income').interactive()\n", "\n", "# Save the chart\n", "chart.save('spending_score_vs_annual_income_scatter_plot.json')\n", "chart" ], "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "\n", "\n", "
\n", "" ], "text/plain": [ "alt.Chart(...)" ] }, "metadata": {}, "execution_count": 11 } ], "execution_count": null, "metadata": { "id": "vr0n7L5eXGC7", "colab": { "base_uri": "https://localhost:8080/", "height": 385 }, "outputId": "79d9a218-83ea-4daa-bafa-261fd71edcd1" } }, { "cell_type": "markdown", "source": [ "[json-tag: code-generated-json-7-0-1738159089351249376]\n", "\n", "I'll now demonstrate how to determine the optimal values for the `eps` and `min_samples` hyperparameters in DBSCAN. These hyperparameters are crucial for controlling the density of clusters.\n", "\n", "I'll use the k-nearest neighbors (KNN) algorithm to help find a suitable value for `eps`. The KNN algorithm calculates the distances between data points, which can be useful in identifying areas of high density in the dataset." ], "metadata": { "id": "limMjxqEXGC7" } }, { "cell_type": "code", "source": [ "from sklearn.cluster import DBSCAN\n", "\n", "# Create a DBSCAN model\n", "dbscan = DBSCAN(eps=9, min_samples=5)\n", "\n", "# Fit the model and predict cluster labels\n", "X['Cluster'] = dbscan.fit_predict(X)\n", "\n", "# Count the number of data points in each cluster\n", "cluster_counts = X['Cluster'].value_counts()\n", "print(cluster_counts.to_markdown(numalign=\"left\", stralign=\"left\"))\n", "\n", "# Create a scatter plot\n", "chart = alt.Chart(X).mark_point().encode(\n", " x='Annual Income (k$)',\n", " y='Spending Score (1-100)',\n", " color='Cluster:N',\n", " tooltip=['Annual Income (k$)', 'Spending Score (1-100)', 'Cluster']\n", ").properties(title='DBSCAN Clustering with eps=9 and min_samples=5').interactive()\n", "\n", "# Save the chart\n", "chart.save('dbscan_clustering_scatter_plot.json')\n", "chart" ], "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "| Cluster | count |\n", "|:----------|:--------|\n", "| 3 | 88 |\n", "| 4 | 31 |\n", "| -1 | 23 |\n", "| 5 | 23 |\n", "| 0 | 16 |\n", "| 1 | 12 |\n", "| 2 | 7 |\n" ] }, { "output_type": "stream", "name": "stderr", "text": [ ":7: SettingWithCopyWarning: \n", "A value is trying to be set on a copy of a slice from a DataFrame.\n", "Try using .loc[row_indexer,col_indexer] = value instead\n", "\n", "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n", " X['Cluster'] = dbscan.fit_predict(X)\n" ] }, { "output_type": "execute_result", "data": { "text/html": [ "\n", "\n", "
\n", "" ], "text/plain": [ "alt.Chart(...)" ] }, "metadata": {}, "execution_count": 12 } ], "execution_count": null, "metadata": { "id": "OjEG8dS4XGC8", "colab": { "base_uri": "https://localhost:8080/", "height": 657 }, "outputId": "fac9b86d-80b9-403a-d15a-26b5ea4f4cdc" } }, { "cell_type": "code", "source": [], "metadata": { "id": "GUQeNtccM20l" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "source": [], "metadata": { "id": "OB_V5PGzM2pT" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "source": [ "\n", "\n", "Here is a step-by-step code for explaining DBSCAN:" ], "metadata": { "id": "uNnLNxTGXGC8" } }, { "cell_type": "code", "source": [ "import pandas as pd\n", "import altair as alt\n", "from sklearn.neighbors import NearestNeighbors\n", "from sklearn.cluster import DBSCAN\n", "\n", "pd.set_option('display.max_rows', None)\n", "pd.set_option('display.max_columns', None)\n", "\n", "# Read the CSV file into a DataFrame\n", "df = pd.read_csv('Mall_Customers.csv')\n", "\n", "# Display the first 5 rows\n", "print(df.head().to_markdown(index=False, numalign=\"left\", stralign=\"left\"))\n", "\n", "# Print the column names and their data types\n", "print(df.info())\n", "\n", "# Extract the features\n", "X = df[['Annual Income (k$)', 'Spending Score (1-100)']]\n", "\n", "# Create a scatter plot\n", "chart = alt.Chart(X).mark_point().encode(\n", " x='Annual Income (k$)',\n", " y='Spending Score (1-100)',\n", " tooltip=['Annual Income (k$)', 'Spending Score (1-100)']\n", ").properties(title='Scatter Plot of Spending Score vs. Annual Income').interactive()\n", "\n", "# Save the chart\n", "chart.save('spending_score_vs_annual_income_scatter_plot.json')\n", "\n", "# Fit a NearestNeighbors model\n", "knn = NearestNeighbors(n_neighbors=5)\n", "knn.fit(X)\n", "\n", "# Calculate distances to the 5th nearest neighbor\n", "distances, indices = knn.kneighbors(X)\n", "distances = np.sort(distances[:, 4])\n", "\n", "# Create a DataFrame for plotting\n", "df_distances = pd.DataFrame({'Distance': distances, 'Point': range(1, len(distances) + 1)})\n", "\n", "# Create a line chart\n", "chart = alt.Chart(df_distances).mark_line(point=True).encode(\n", " x='Point',\n", " y='Distance',\n", " tooltip=['Point', 'Distance']\n", ").properties(title='Sorted Distances to the 5th Nearest Neighbor').interactive()\n", "\n", "# Save the chart\n", "chart.save('sorted_distances_5th_nearest_neighbor.json')\n", "\n", "# Create a DBSCAN model\n", "dbscan = DBSCAN(eps=9, min_samples=5)\n", "\n", "# Fit the model and predict cluster labels\n", "X['Cluster'] = dbscan.fit_predict(X)\n", "\n", "# Count the number of data points in each cluster\n", "cluster_counts = X['Cluster'].value_counts()\n", "print(cluster_counts.to_markdown(numalign=\"left\", stralign=\"left\"))\n", "\n", "# Create a scatter plot\n", "chart = alt.Chart(X).mark_point().encode(\n", " x='Annual Income (k$)',\n", " y='Spending Score (1-100)',\n", " color='Cluster:N',\n", " tooltip=['Annual Income (k$)', 'Spending Score (1-100)', 'Cluster']\n", ").properties(title='DBSCAN Clustering with eps=9 and min_samples=5').interactive()\n", "\n", "# Save the chart\n", "chart.save('dbscan_clustering_scatter_plot.json')" ], "outputs": [], "execution_count": null, "metadata": { "id": "vmFkEnxgXGC8" } } ], "metadata": { "colab": { "provenance": [] }, "kernelspec": { "display_name": "Python 3", "name": "python3" } }, "nbformat": 4, "nbformat_minor": 0 }