How to Build a Municipal Crime Data Dashboard with Python

#python #tutorial #webdev #programming

Cities publish crime data through open data portals in hundreds of different formats. Build a unified dashboard that aggregates and visualizes crime trends across municipalities.

The Problem with Municipal Data

Every city publishes crime data differently — Chicago uses CSV, NYC uses an API, LA uses a data portal. A unified scraper normalizes everything into one comparable format.

Scraping City Crime Portals

import requests
import pandas as pd

API_KEY = "YOUR_SCRAPERAPI_KEY"  # Get one at https://www.scraperapi.com?fp_ref=the52

def scrape_chicago_crimes(limit=1000):
    url = f"https://data.cityofchicago.org/resource/ijzp-q8t2.json?$limit={limit}&$order=date DESC"
    response = requests.get(url, timeout=30)
    data = response.json()

    records = []
    for crime in data:
        records.append({
            'city': 'Chicago',
            'type': crime.get('primary_type', ''),
            'date': crime.get('date', ''),
            'latitude': crime.get('latitude'),
            'longitude': crime.get('longitude'),
            'location': crime.get('block', '')
        })
    return pd.DataFrame(records)

def scrape_nyc_crimes(limit=1000):
    url = f"https://data.cityofnewyork.us/resource/5uac-w243.json?$limit={limit}&$order=cmplnt_fr_dt DESC"
    response = requests.get(url, timeout=30)
    data = response.json()

    records = []
    for crime in data:
        records.append({
            'city': 'NYC',
            'type': crime.get('ofns_desc', ''),
            'date': crime.get('cmplnt_fr_dt', ''),
            'latitude': crime.get('latitude'),
            'longitude': crime.get('longitude'),
            'location': crime.get('boro_nm', '')
        })
    return pd.DataFrame(records)

chicago = scrape_chicago_crimes()
nyc = scrape_nyc_crimes()
combined = pd.concat([chicago, nyc], ignore_index=True)
print(f"Total records: {len(combined)}")

Adding More Cities

def scrape_la_crimes():
    url = "https://data.lacity.org/resource/2nrs-mtv8.json?$limit=1000&$order=date_occ DESC"
    response = requests.get(url, timeout=30)
    data = response.json()

    records = []
    for crime in data:
        records.append({
            'city': 'Los Angeles',
            'type': crime.get('crm_cd_desc', ''),
            'date': crime.get('date_occ', ''),
            'latitude': crime.get('lat'),
            'longitude': crime.get('lon'),
            'location': crime.get('area_name', '')
        })
    return pd.DataFrame(records)

la = scrape_la_crimes()
combined = pd.concat([combined, la], ignore_index=True)

Analyzing Crime Patterns

import matplotlib.pyplot as plt

combined['date_parsed'] = pd.to_datetime(combined['date'], errors='coerce')
combined['month'] = combined['date_parsed'].dt.to_period('M')

monthly = combined.groupby(['city', 'month']).size().reset_index(name='count')
for city in monthly['city'].unique():
    city_data = monthly[monthly['city'] == city]
    plt.plot(city_data['month'].astype(str), city_data['count'], label=city)

plt.legend()
plt.title('Monthly Crime Reports by City')
plt.xticks(rotation=45)
plt.tight_layout()
plt.savefig('crime_trends.png', dpi=150)

Building the Flask Dashboard

Use ThorData for government portals. Track with ScrapeOps:

from flask import Flask, render_template_string

app = Flask(__name__)

@app.route('/')
def dashboard():
    stats = combined.groupby('city').agg(
        total=('type', 'count'),
        most_common=('type', lambda x: x.mode().iloc[0] if len(x) > 0 else 'N/A')
    ).to_dict('index')

    return render_template_string('''
    <h1>Crime Data Dashboard</h1>
    {% for city, data in stats.items() %}
    <h2>{{ city }}</h2>
    <p>Total: {{ data.total }} | Most common: {{ data.most_common }}</p>
    {% endfor %}
    ''', stats=stats)

if __name__ == '__main__':
    app.run(debug=True, port=5000)