File size: 17,798 Bytes
46e47b6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
{
  "cells": [
    {
      "cell_type": "markdown",
      "id": "b1c6a137",
      "metadata": {
        "id": "b1c6a137"
      },
      "source": [
        "# Clustering Lab: State Crime Pattern Analysis\n",
        "\n",
        "## Lab Overview\n",
        "\n",
        "Welcome to your hands-on clustering lab! You'll be working as a policy analyst for the Department of Justice, analyzing crime patterns across US states. Your mission: discover hidden safety profiles that could inform federal resource allocation and crime prevention strategies.\n",
        "\n",
        "**Your Deliverable**: A policy brief with visualizations and recommendations based on your clustering analysis.\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 1: Data Detective Work\n",
        "**Time: 15 minutes | Product: Data Summary Report**\n",
        "\n",
        "### Your Task\n",
        "Before any analysis, you need to understand what you're working with. Create a brief data summary that a non-technical policy maker could understand.\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from statsmodels.datasets import get_rdataset\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.cluster import KMeans, AgglomerativeClustering\n",
        "\n",
        "# Load the data\n",
        "USArrests = get_rdataset('USArrests').data\n",
        "print(\"Dataset shape:\", USArrests.shape)\n",
        "print(\"\\nVariables:\", USArrests.columns.tolist())\n",
        "print(\"\\nFirst 5 states:\")\n",
        "print(USArrests.head())\n",
        "```"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 106
        },
        "id": "mqRVE1hlXK9x",
        "outputId": "5a1bbd64-15cd-4e1c-9344-64a901d8a396"
      },
      "id": "mqRVE1hlXK9x",
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "SyntaxError",
          "evalue": "invalid syntax (<ipython-input-1-2035427107>, line 1)",
          "traceback": [
            "\u001b[0;36m  File \u001b[0;32m\"<ipython-input-1-2035427107>\"\u001b[0;36m, line \u001b[0;32m1\u001b[0m\n\u001b[0;31m    ```python\u001b[0m\n\u001b[0m    ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Your Investigation\n",
        "Complete this data summary table:\n",
        "\n",
        "| Variable | What it measures | Average Value | Highest State | Lowest State |\n",
        "|----------|------------------|---------------|---------------|--------------|\n",
        "| Murder | Rate per 100,000 people | ??? | ??? | ??? |\n",
        "| Assault | Rate per 100,000 people | ??? | ??? | ??? |\n",
        "| UrbanPop | Percentage living in cities | ??? | ??? | ??? |\n",
        "| Rape | Rate per 100,000 people | ??? | ??? | ??? |\n",
        "\n",
        "**Deliverable**: Write 2-3 sentences describing the biggest surprises in this data. Which states are not what you expected?\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 2: The Scaling Challenge\n",
        "**Time: 10 minutes | Product: Before/After Comparison**\n",
        "\n",
        "### Your Task\n",
        "Demonstrate why scaling is critical for clustering crime data.\n",
        "\n"
      ],
      "metadata": {
        "id": "7qkDKTe4XLtG"
      },
      "id": "7qkDKTe4XLtG"
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "# Check the scale differences\n",
        "print(\"Original data ranges:\")\n",
        "print(USArrests.describe())\n",
        "\n",
        "print(\"\\nVariances (how spread out the data is):\")\n",
        "print(USArrests.var())\n",
        "\n",
        "# Scale the data\n",
        "scaler = StandardScaler()\n",
        "USArrests_scaled = scaler.fit_transform(USArrests)\n",
        "scaled_df = pd.DataFrame(USArrests_scaled,\n",
        "                        columns=USArrests.columns,\n",
        "                        index=USArrests.index)\n",
        "\n",
        "print(\"\\nAfter scaling - all variables now have similar ranges:\")\n",
        "print(scaled_df.describe())\n",
        "```"
      ],
      "metadata": {
        "id": "zQ3VowYNXLeQ"
      },
      "id": "zQ3VowYNXLeQ",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Your Analysis\n",
        "1. **Before scaling**: Which variable would dominate the clustering? Why?\n",
        "2. **After scaling**: Explain in simple terms what StandardScaler did to the data.\n",
        "\n",
        "**Deliverable**: One paragraph explaining why a policy analyst should care about data scaling.\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 3: Finding the Right Number of Groups\n",
        "**Time: 20 minutes | Product: Recommendation with Visual Evidence**\n",
        "\n",
        "### Your Task\n",
        "Use the elbow method to determine how many distinct crime profiles exist among US states.\n"
      ],
      "metadata": {
        "id": "FnOT700SXLPh"
      },
      "id": "FnOT700SXLPh"
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "# Test different numbers of clusters\n",
        "inertias = []\n",
        "K_values = range(1, 11)\n",
        "\n",
        "for k in K_values:\n",
        "    kmeans = KMeans(n_clusters=k, random_state=42, n_init=20)\n",
        "    kmeans.fit(USArrests_scaled)\n",
        "    inertias.append(kmeans.inertia_)\n",
        "\n",
        "# Create the elbow plot\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(K_values, inertias, 'bo-', linewidth=2, markersize=8)\n",
        "plt.xlabel('Number of Clusters (K)')\n",
        "plt.ylabel('Within-Cluster Sum of Squares')\n",
        "plt.title('Finding the Optimal Number of State Crime Profiles')\n",
        "plt.grid(True, alpha=0.3)\n",
        "plt.show()\n",
        "\n",
        "# Print the inertia values\n",
        "for k, inertia in zip(K_values, inertias):\n",
        "    print(f\"K={k}: Inertia = {inertia:.1f}\")\n",
        "```"
      ],
      "metadata": {
        "id": "zOQrS9lmXpTF"
      },
      "id": "zOQrS9lmXpTF",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "id": "2e388ef2",
      "metadata": {
        "id": "2e388ef2"
      },
      "source": [
        "### Your Decision\n",
        "Based on your elbow plot:\n",
        "1. **What value of K do you recommend?** (Look for the \"elbow\" where the line starts to flatten)\n",
        "2. **What does this mean in policy terms?** (How many distinct types of state crime profiles exist?)\n",
        "\n",
        "**Deliverable**: A one-paragraph recommendation with your chosen K value and reasoning.\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 4: K-Means State Profiling\n",
        "**Time: 25 minutes | Product: State Crime Profile Report**\n",
        "\n",
        "### Your Task\n",
        "Create distinct crime profiles and identify which states belong to each category.\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "# Use your chosen K value from Exercise 3\n",
        "optimal_k = 4  # Replace with your chosen value\n",
        "\n",
        "# Perform K-means clustering\n",
        "kmeans = KMeans(n_clusters=optimal_k, random_state=42, n_init=20)\n",
        "cluster_labels = kmeans.fit_predict(USArrests_scaled)\n",
        "\n",
        "# Add cluster labels to original data\n",
        "USArrests_clustered = USArrests.copy()\n",
        "USArrests_clustered['Cluster'] = cluster_labels\n",
        "\n",
        "# Analyze each cluster\n",
        "print(\"State Crime Profiles Analysis\")\n",
        "print(\"=\" * 50)\n",
        "\n",
        "for cluster_num in range(optimal_k):\n",
        "    cluster_states = USArrests_clustered[USArrests_clustered['Cluster'] == cluster_num]\n",
        "    print(f\"\\nCLUSTER {cluster_num}: {len(cluster_states)} states\")\n",
        "    print(\"States:\", \", \".join(cluster_states.index.tolist()))\n",
        "    print(\"Average characteristics:\")\n",
        "    avg_profile = cluster_states[['Murder', 'Assault', 'UrbanPop', 'Rape']].mean()\n",
        "    for var, value in avg_profile.items():\n",
        "        print(f\"  {var}: {value:.1f}\")\n",
        "```"
      ],
      "metadata": {
        "id": "_5b0nE6KXv1P"
      },
      "id": "_5b0nE6KXv1P",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Your Analysis\n",
        "For each cluster, create a profile:\n",
        "\n",
        "**Cluster 0: \"[Your Creative Name]\"**\n",
        "- **States**: [List them]\n",
        "- **Characteristics**: [Describe the pattern]\n",
        "- **Policy Insight**: [What should federal agencies know about these states?]\n",
        "\n",
        "**Deliverable**: A table summarizing each cluster with creative names and policy recommendations.\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 5: Hierarchical Clustering Exploration\n",
        "**Time: 25 minutes | Product: Family Tree Interpretation**\n",
        "\n",
        "### Your Task\n",
        "Create a dendrogram to understand how states naturally group together.\n"
      ],
      "metadata": {
        "id": "J1WVGb_nX4ye"
      },
      "id": "J1WVGb_nX4ye"
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "from scipy.cluster.hierarchy import dendrogram, linkage\n",
        "\n",
        "# Create hierarchical clustering\n",
        "linkage_matrix = linkage(USArrests_scaled, method='complete')\n",
        "\n",
        "# Plot the dendrogram\n",
        "plt.figure(figsize=(15, 8))\n",
        "dendrogram(linkage_matrix,\n",
        "           labels=USArrests.index.tolist(),\n",
        "           leaf_rotation=90,\n",
        "           leaf_font_size=10)\n",
        "plt.title('State Crime Pattern Family Tree')\n",
        "plt.xlabel('States')\n",
        "plt.ylabel('Distance Between Groups')\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "```"
      ],
      "metadata": {
        "id": "Y9a_cbZKX7QX"
      },
      "id": "Y9a_cbZKX7QX",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Your Interpretation\n",
        "1. **Closest Pairs**: Which two states are most similar in crime patterns?\n",
        "2. **Biggest Divide**: Where is the largest split in the tree? What does this represent?\n",
        "3. **Surprising Neighbors**: Which states cluster together that surprised you geographically?\n",
        "\n",
        "### Code to Compare Methods"
      ],
      "metadata": {
        "id": "0PaImqZtX6f3"
      },
      "id": "0PaImqZtX6f3"
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "# Compare your K-means results with hierarchical clustering\n",
        "from scipy.cluster.hierarchy import fcluster\n",
        "\n",
        "# Cut the tree to get the same number of clusters as K-means\n",
        "hierarchical_labels = fcluster(linkage_matrix, optimal_k, criterion='maxclust') - 1\n",
        "\n",
        "# Create comparison\n",
        "comparison_df = pd.DataFrame({\n",
        "    'State': USArrests.index,\n",
        "    'K_Means_Cluster': cluster_labels,\n",
        "    'Hierarchical_Cluster': hierarchical_labels\n",
        "})\n",
        "\n",
        "print(\"Comparison of K-Means vs Hierarchical Clustering:\")\n",
        "print(comparison_df.sort_values('State'))\n",
        "\n",
        "# Count agreements\n",
        "agreements = sum(comparison_df['K_Means_Cluster'] == comparison_df['Hierarchical_Cluster'])\n",
        "print(f\"\\nMethods agreed on {agreements} out of {len(comparison_df)} states ({agreements/len(comparison_df)*100:.1f}%)\")\n",
        "```"
      ],
      "metadata": {
        "id": "tJQ-C5GFYBRT"
      },
      "id": "tJQ-C5GFYBRT",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Deliverable**: A paragraph explaining the key differences between what K-means and hierarchical clustering revealed.\n",
        "\n",
        "---\n",
        "\n",
        "## Exercise 6: Policy Brief Creation\n",
        "**Time: 20 minutes | Product: Executive Summary**\n",
        "\n",
        "### Your Task\n",
        "Synthesize your findings into a policy brief for Department of Justice leadership.\n",
        "\n",
        "### Code Framework for Final Visualization"
      ],
      "metadata": {
        "id": "dx1fNhu4YD7-"
      },
      "id": "dx1fNhu4YD7-"
    },
    {
      "cell_type": "code",
      "source": [
        "```python\n",
        "# Create a comprehensive visualization\n",
        "fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(15, 12))\n",
        "\n",
        "# Plot 1: Murder vs Assault by cluster\n",
        "colors = ['red', 'blue', 'green', 'orange', 'purple']\n",
        "for i in range(optimal_k):\n",
        "    cluster_data = USArrests_clustered[USArrests_clustered['Cluster'] == i]\n",
        "    ax1.scatter(cluster_data['Murder'], cluster_data['Assault'],\n",
        "               c=colors[i], label=f'Cluster {i}', s=60, alpha=0.7)\n",
        "ax1.set_xlabel('Murder Rate')\n",
        "ax1.set_ylabel('Assault Rate')\n",
        "ax1.set_title('Murder vs Assault by Crime Profile')\n",
        "ax1.legend()\n",
        "ax1.grid(True, alpha=0.3)\n",
        "\n",
        "# Plot 2: Urban Population vs Rape by cluster\n",
        "for i in range(optimal_k):\n",
        "    cluster_data = USArrests_clustered[USArrests_clustered['Cluster'] == i]\n",
        "    ax2.scatter(cluster_data['UrbanPop'], cluster_data['Rape'],\n",
        "               c=colors[i], label=f'Cluster {i}', s=60, alpha=0.7)\n",
        "ax2.set_xlabel('Urban Population %')\n",
        "ax2.set_ylabel('Rape Rate')\n",
        "ax2.set_title('Urban Population vs Rape Rate by Crime Profile')\n",
        "ax2.legend()\n",
        "ax2.grid(True, alpha=0.3)\n",
        "\n",
        "# Plot 3: Cluster size comparison\n",
        "cluster_sizes = USArrests_clustered['Cluster'].value_counts().sort_index()\n",
        "ax3.bar(range(len(cluster_sizes)), cluster_sizes.values, color=colors[:len(cluster_sizes)])\n",
        "ax3.set_xlabel('Cluster Number')\n",
        "ax3.set_ylabel('Number of States')\n",
        "ax3.set_title('Number of States in Each Crime Profile')\n",
        "ax3.set_xticks(range(len(cluster_sizes)))\n",
        "\n",
        "# Plot 4: Average crime rates by cluster\n",
        "cluster_means = USArrests_clustered.groupby('Cluster')[['Murder', 'Assault', 'Rape']].mean()\n",
        "cluster_means.plot(kind='bar', ax=ax4)\n",
        "ax4.set_xlabel('Cluster Number')\n",
        "ax4.set_ylabel('Average Rate')\n",
        "ax4.set_title('Average Crime Rates by Profile')\n",
        "ax4.legend()\n",
        "ax4.tick_params(axis='x', rotation=0)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "```"
      ],
      "metadata": {
        "id": "N8bkxURpYHJF"
      },
      "id": "N8bkxURpYHJF",
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Your Policy Brief Template\n",
        "\n",
        "**EXECUTIVE SUMMARY: US State Crime Profile Analysis**\n",
        "\n",
        "**Key Findings:**\n",
        "- We identified [X] distinct crime profiles among US states\n",
        "- [State examples] represent the highest-risk profile\n",
        "- [State examples] represent the lowest-risk profile\n",
        "- Urban population [does/does not] strongly correlate with violent crime\n",
        "\n",
        "**Policy Recommendations:**\n",
        "1. **High-Priority States**: [List and explain why]\n",
        "2. **Resource Allocation**: [Suggest how to distribute federal crime prevention funds]\n",
        "3. **Best Practice Sharing**: [Which states should learn from which others?]\n",
        "\n",
        "**Methodology Note**: Analysis used unsupervised clustering on 4 crime variables across 50 states, with data standardization to ensure fair comparison.\n",
        "\n",
        "**Deliverable**: A complete 1-page policy brief with your clustering insights and specific recommendations.\n"
      ],
      "metadata": {
        "id": "rAy_Ye0WYLK0"
      },
      "id": "rAy_Ye0WYLK0"
    }
  ],
  "metadata": {
    "jupytext": {
      "cell_metadata_filter": "-all",
      "formats": "Rmd,ipynb",
      "main_language": "python"
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.4"
    },
    "colab": {
      "provenance": []
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}