Visualising SimPy Results: Making Data Speak

Published on February 07, 2026

Numbers tell the story. Visualisations make people listen. <h2 id="essential-plots">Essential Plots</h2> Every simulation should produce: <ol> <li>Queue length over time - System behaviour</li> <li>Wait time distribution - Customer experience</li> <li>Utilisation over time - Resource efficiency</li> <li>Throughput chart - System output</li> </ol> <h2 id="setup">Setup</h2> <div class="code-block"><pre><code>import simpy import random import pandas as pd import matplotlib.pyplot as plt import numpy as np # Style plt.style.use('seaborn-v0_8-whitegrid') plt.rcParams['figure.figsize'] = (10, 6) </code></pre></div> <h2 id="queue-length-over-time">Queue Length Over Time</h2> <div class="code-block"><pre><code>def plot_queue_length(time_series_df): plt.figure() plt.step(time_series_df['time'], time_series_df['queue_length'], where='post', linewidth=1) plt.fill_between(time_series_df['time'], time_series_df['queue_length'], step='post', alpha=0.3) plt.xlabel('Time') plt.ylabel('Queue Length') plt.title('Queue Length Over Time') plt.tight_layout() plt.savefig('queue_length.png', dpi=150) plt.show() </code></pre></div> <h2 id="wait-time-histogram">Wait Time Histogram</h2> <div class="code-block"><pre><code>def plot_wait_distribution(wait_times): plt.figure() plt.hist(wait_times, bins=30, edgecolor='white', alpha=0.7) plt.axvline(np.mean(wait_times), color='red', linestyle='--', label=f'Mean: {np.mean(wait_times):.2f}') plt.axvline(np.percentile(wait_times, 95), color='orange', linestyle='--', label=f'95th percentile: {np.percentile(wait_times, 95):.2f}') plt.xlabel('Wait Time') plt.ylabel('Frequency') plt.title('Distribution of Wait Times') plt.legend() plt.tight_layout() plt.savefig('wait_distribution.png', dpi=150) plt.show() </code></pre></div> <h2 id="utilisation-chart">Utilisation Chart</h2> <div class="code-block"><pre><code>def plot_utilisation(time_series_df, resource_name='server'): plt.figure() plt.plot(time_series_df['time'], time_series_df[f'{resource_name}_utilisation'], linewidth=1) plt.fill_between(time_series_df['time'], time_series_df[f'{resource_name}_utilisation'], alpha=0.3) avg_util = time_series_df[f'{resource_name}_utilisation'].mean() plt.axhline(avg_util, color='red', linestyle='--', label=f'Average: {avg_util:.1%}') plt.xlabel('Time') plt.ylabel('Utilisation') plt.ylim(0, 1.1) plt.title(f'{resource_name.title()} Utilisation Over Time') plt.legend() plt.tight_layout() plt.savefig('utilisation.png', dpi=150) plt.show() </code></pre></div> <h2 id="box-plot-comparison">Box Plot Comparison</h2> Compare scenarios: <div class="code-block"><pre><code>def plot_scenario_comparison(results_dict): """results_dict = {'Scenario A': wait_times_a, 'Scenario B': wait_times_b}""" plt.figure() data = list(results_dict.values()) labels = list(results_dict.keys()) plt.boxplot(data, labels=labels) plt.ylabel('Wait Time') plt.title('Wait Time Comparison Across Scenarios') plt.tight_layout() plt.savefig('comparison.png', dpi=150) plt.show() </code></pre></div> <h2 id="throughput-over-time">Throughput Over Time</h2> <div class="code-block"><pre><code>def plot_throughput(entity_df, window=50): """Cumulative completions and rolling throughput.""" fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 8)) # Cumulative completions entity_df = entity_df.sort_values('departure') entity_df['cumulative'] = range(1, len(entity_df) + 1) ax1.plot(entity_df['departure'], entity_df['cumulative']) ax1.set_xlabel('Time') ax1.set_ylabel('Cumulative Completions') ax1.set_title('Cumulative Throughput') # Rolling throughput entity_df['throughput'] = entity_df['cumulative'] / entity_df['departure'] ax2.plot(entity_df['departure'], entity_df['throughput']) ax2.set_xlabel('Time') ax2.set_ylabel('Throughput (per time unit)') ax2.set_title('Throughput Rate') plt.tight_layout() plt.savefig('throughput.png', dpi=150) plt.show() </code></pre></div> <h2 id="multiple-resources">Multiple Resources</h2> <div class="code-block"><pre><code>def plot_multiple_resources(time_series_df, resources): fig, axes = plt.subplots(len(resources), 1, figsize=(10, 4*len(resources))) for ax, resource in zip(axes, resources): ax.step(time_series_df['time'], time_series_df[f'{resource}_queue'], where='post', label='Queue') ax.step(time_series_df['time'], time_series_df[f'{resource}_busy'], where='post', label='In Service') ax.set_xlabel('Time') ax.set_ylabel('Count') ax.set_title(f'{resource.title()}') ax.legend() plt.tight_layout() plt.savefig('resources.png', dpi=150) plt.show() </code></pre></div> <h2 id="heatmap-of-arrivals">Heatmap of Arrivals</h2> <div class="code-block"><pre><code>def plot_arrival_heatmap(entity_df, bins_x=24, bins_y=7): """Show arrival patterns (e.g., hour of day vs day of week).""" # Assuming arrival times can be converted to hour/day entity_df['hour'] = (entity_df['arrival'] % 24).astype(int) entity_df['day'] = (entity_df['arrival'] // 24 % 7).astype(int) heatmap_data = entity_df.groupby(['day', 'hour']).size().unstack(fill_value=0) plt.figure(figsize=(12, 5)) plt.imshow(heatmap_data, aspect='auto', cmap='YlOrRd') plt.colorbar(label='Arrivals') plt.xlabel('Hour of Day') plt.ylabel('Day of Week') plt.title('Arrival Pattern Heatmap') plt.tight_layout() plt.savefig('heatmap.png', dpi=150) plt.show() </code></pre></div> <h2 id="cdf-plot">CDF Plot</h2> <div class="code-block"><pre><code>def plot_cdf(wait_times, label='Wait Time'): sorted_data = np.sort(wait_times) cdf = np.arange(1, len(sorted_data) + 1) / len(sorted_data) plt.figure() plt.plot(sorted_data, cdf, linewidth=2) plt.xlabel(label) plt.ylabel('Cumulative Probability') plt.title(f'Cumulative Distribution of {label}') # Mark key percentiles for p in [50, 90, 95, 99]: val = np.percentile(wait_times, p) plt.axhline(p/100, color='gray', linestyle=':', alpha=0.5) plt.axvline(val, color='gray', linestyle=':', alpha=0.5) plt.annotate(f'{p}%: {val:.1f}', xy=(val, p/100), xytext=(val + 1, p/100 - 0.05)) plt.tight_layout() plt.savefig('cdf.png', dpi=150) plt.show() </code></pre></div> <h2 id="confidence-interval-plot">Confidence Interval Plot</h2> <div class="code-block"><pre><code>def plot_confidence_intervals(replication_results): """replication_results = list of mean wait times from each replication.""" n = len(replication_results) mean = np.mean(replication_results) std = np.std(replication_results, ddof=1) ci = 1.96 * std / np.sqrt(n) plt.figure() plt.bar(['Mean Wait Time'], [mean], yerr=[ci], capsize=10, color='steelblue') plt.ylabel('Time') plt.title(f'Mean Wait Time with 95% CI\n(n={n} replications)') plt.tight_layout() plt.savefig('confidence.png', dpi=150) plt.show() </code></pre></div> <h2 id="interactive-with-plotly">Interactive with Plotly</h2> For exploration: <div class="code-block"><pre><code>import plotly.express as px import plotly.graph_objects as go def interactive_queue(time_series_df): fig = px.line(time_series_df, x='time', y='queue_length', title='Queue Length Over Time') fig.update_traces(line_shape='hv') # Step plot fig.write_html('queue_interactive.html') fig.show() </code></pre></div> <h2 id="complete-visualisation-suite">Complete Visualisation Suite</h2> <div class="code-block"><pre><code>def generate_all_plots(entity_df, time_series_df, resources): """Generate all standard visualisations.""" # 1. Queue length for resource in resources: plot_queue_length(time_series_df.rename( columns={f'{resource}_queue': 'queue_length'} )) # 2. Wait distribution plot_wait_distribution(entity_df['wait'].values) # 3. Utilisation for resource in resources: plot_utilisation(time_series_df, resource) # 4. CDF plot_cdf(entity_df['wait'].values) # 5. Throughput plot_throughput(entity_df) print("All plots generated!") </code></pre></div> <h2 id="summary">Summary</h2> Good visualisations: - Show trends over time (not just averages) - Include distributions (not just means) - Compare scenarios side-by-side - Mark key thresholds and targets Numbers inform. Pictures convince. <h2 id="next-steps">Next Steps</h2> <ul> <li><a href="/blog_posts/simpy-collect-statistics.html">Collecting Statistics</a></li> <li><a href="/blog_posts/simpy-run-multiple-replications.html">Running Multiple Replications</a></li> <li><a href="/blog_posts/simpy-with-pandas.html">SimPy with pandas</a></li> </ul>

Discover the Power of Simulation

Want to become a go-to expert in simulation with Python? The Complete Simulation Bootcamp will show you how simulation can transform your career and your projects.

Explore the Bootcamp