data-visualization

# Data Visualization ## What I do I provide tools and techniques for creating effective visual representations of data. I enable you to explore data patterns through exploratory plots, communicate insights through explanatory visualizations, and build interactive dashboards. Good visualization helps stakeholders understand complex data quickly and supports data-driven decision making. ## When to use me - Exploring data distributions and relationships - Identifying patterns, trends, and anomalies - Comparing groups or categories - Showing changes over time - Presenting findings to stakeholders - Building interactive dashboards - Reporting analysis results - Communicating uncertainty ## Core Concepts ### Chart Types - **Distribution**: Histogram, KDE, box plot, violin plot - **Relationship**: Scatter plot, line plot, heatmap - **Comparison**: Bar chart, grouped bar chart, bubble chart - **Composition**: Pie chart, stacked bar, treemap - **Time Series**: Line chart, area chart, candlestick ### Design Principles - **Clarity**: Clear titles, labels, and legends - **Simplicity**: Avoid chart junk and unnecessary elements - **Color**: Use appropriate color schemes (sequential, diverging, qualitative) - **Scale**: Use appropriate axis scales (linear, log) - **Context**: Include reference lines, annotations, and context ### Tools - **Matplotlib**: Low-level, flexible, publication-quality plots - **Seaborn**: High-level statistical visualizations - **Plotly**: Interactive plots and dashboards - **Altair**: Declarative visualization ## Code Examples (Python) ```python import matplotlib.pyplot as plt import seaborn as sns import numpy as np import pandas as pd # Set style plt.style.use('seaborn-v0_8-whitegrid') sns.set_palette("husl") # Basic plots with matplotlib fig, axes = plt.subplots(2, 2, figsize=(12, 10)) # Line plot axes[0, 0].plot(x, y, 'b-', linewidth=2, marker='o', markersize=4) axes[0, 0].set_xlabel('X Axis') axes[0, 0].set_ylabel('Y Axis') axes[0, 0].set_title('Line Plot') # Scatter plot axes[0, 1].scatter(x, y, c=z, cmap='viridis', alpha=0.7, s=50) axes[0, 1].set_xlabel('X Axis') axes[0, 1].set_ylabel('Y Axis') axes[0, 1].set_title('Scatter Plot') # Bar chart axes[1, 0].bar(categories, values, color=['#1f77b4', '#ff7f0e', '#2ca02c']) axes[1, 0].set_xlabel('Category') axes[1, 0].set_ylabel('Value') axes[1, 0].set_title('Bar Chart') # Histogram axes[1, 1].hist(data, bins=30, color='steelblue', edgecolor='white', alpha=0.7) axes[1, 1].set_xlabel('Value') axes[1, 1].set_ylabel('Frequency') axes[1, 1].set_title('Histogram') plt.tight_layout() plt.savefig('basic_plots.png', dpi=150, bbox_inches='tight') plt.show() # Seaborn plots # Distribution plots fig, axes = plt.subplots(2, 2, figsize=(12, 10)) # Histogram with KDE sns.histplot(data, kde=True, ax=axes[0, 0], color='skyblue') # Box plot sns.boxplot(x='category', y='value', data=df, ax=axes[0, 1]) # Violin plot sns.violinplot(x='category', y='value', data=df, ax=axes[1, 0]) # Pair plot for multiple variables # sns.pairplot(df, hue='category', diag_kind='kde') # Relationship plots fig, axes = plt.subplots(1, 3, figsize=(15, 5)) # Scatter with regression line sns.regplot(x='x', y='y', data=df, ax=axes[0]) # Scatter with hue sns.scatterplot(x='x', y='y', hue='category', data=df, ax=axes[1]) # Heatmap correlation_matrix = df.corr() sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', center=0, ax=axes[2]) plt.tight_layout() plt.show() # Time series visualization fig, axes = plt.subplots(2, 1, figsize=(12, 8)) # Line chart with confidence interval sns.lineplot(x='date', y='value', data=df, ax=axes[0]) axes[0].set_title('Time Series') # Area chart sns.lineplot(x='date', y='value', data=df, ax=axes[1], fill=True) axes[1].set_title('Area Chart') plt.tight_layout() plt.show() # Categorical data visualization fig, axes = plt.subplots(1, 3, figsize=(15, 5)) # Count plot sns.countplot(x='category', data=df, ax=axes[0]) # Bar plot with error bars sns.barplot(x='category', y='value', data=df, ax=axes[1], errorbar='sd') # Grouped bar chart sns.barplot(x='category', y='value', hue='group', data=df, ax=axes[2]) plt.tight_layout() plt.show() # Multi-panel figure fig = plt.figure(figsize=(14, 10)) # GridSpec for custom layout from matplotlib.gridspec import GridSpec gs = GridSpec(3, 3, figure=fig, hspace=0.3, wspace=0.3) # Create subplots ax1 = fig.add_subplot(gs[0, :2]) ax2 = fig.add_subplot(gs[0, 2]) ax3 = fig.add_subplot(gs[1, :]) ax4 = fig.add_subplot(gs[2, 0]) ax5 = fig.add_subplot(gs[2, 1]) ax6 = fig.add_subplot(gs[2, 2]) # Add content ax1.plot(x, y) ax2.bar(categories, values) ax3.scatter(x, y, c=z, cmap='viridis') ax4.hist(data1, bins=20, alpha=0.7) ax5.hist(data2, bins=20, alpha=0.7) ax6.boxplot([data1, data2, data3]) plt.savefig('complex_layout.png', dpi=150, bbox_inches='tight') plt.show() # Annotations and styling fig, ax = plt.subplots(figsize=(10, 6)) ax.plot(x, y, 'b-', linewidth=2) # Add annotations ax.annotate('Peak', xy=(x_max, y_max), xytext=(x_max+5, y_max+10), arrowprops=dict(arrowstyle='->', color='red'), fontsize=12, color='red') # Add reference lines ax.axhline(y=mean_value, color='gray', linestyle='--', label=f'Mean: {mean_value:.2f}') ax.axvline(x=threshold, color='orange', linestyle=':', alpha=0.7) # Add legend ax.legend(loc='upper left') # Styling ax.set_xlabel('X Label', fontsize=12) ax.set_ylabel('Y Label', fontsize=12) ax.set_title('Styled Plot', fontsize=14, fontweight='bold') plt.show() ``` ## Best Practices 1. **Know your audience**: Tailor complexity and detail to the audience's expertise. 2. **Choose the right chart type**: Match the chart to the data and message (comparison vs. distribution vs. relationship). 3. **Keep it simple**: Remove unnecessary elements (chart junk, excessive gridlines, decorative 3D effects). 4. **Use color strategically**: Use color to highlight, not decorate. Use consistent color schemes. 5. **Label clearly**: Axis labels, titles, and legends should be informative and readable. 6. **Provide context**: Include reference points, benchmarks, and relevant annotations. 7. **Consider accessibility**: Use colorblind-friendly palettes and ensure text is readable. 8. **Iterate**: Create multiple versions and get feedback before finalizing. ## Common Patterns ### Pattern 1: Exploratory Data Analysis Dashboard ```python def eda_dashboard(df, numeric_cols, categorical_cols): fig, axes = plt.subplots(len(numeric_cols), 3, figsize=(15, 4*len(numeric_cols))) for i, col in enumerate(numeric_cols): # Distribution sns.histplot(df[col].dropna(), kde=True, ax=axes[i, 0]) axes[i, 0].set_title(f'{col} Distribution') # Box plot by category if categorical_cols: sns.boxplot(x=categorical_cols[0], y=col, data=df, ax=axes[i, 1]) # Outlier summary q1, q3 = df[col].quantile([0.25, 0.75]) iqr = q3 - q1 outliers = df[(df[col] < q1-1.5*iqr) | (df[col] > q3+1.5*iqr)][col].count() axes[i, 2].text(0.5, 0.5, f'Outliers: {outliers}', ha='center', va='center', fontsize=14) axes[i, 2].set_title(f'{col} Summary') plt.tight_layout() return fig ``` ### Pattern 2: Model Performance Comparison ```python def compare_model_performance(results_df): fig, axes = plt.subplots(1, 3, figsize=(15, 5)) # Accuracy comparison sns.barplot(x='model', y='accuracy', data=results_df, ax=axes[0]) axes[0].set_title('Model Accuracy Comparison') axes[0].tick_params(axis='x', rotation=45) # Confusion matrix heatmap sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', ax=axes[1]) axes[1].set_title('Best Model Confusion Matrix') # ROC curves for model_name, fpr, tpr in roc_data: axes[2].plot(fpr, tpr, label=f'{model_name} (AUC={auc:.2f})') axes[2].plot([0, 1], [0, 1], 'k--') axes[2].set_xlabel('False Positive Rate') axes[2].set_ylabel('True Positive Rate') axes[2].set_title('ROC Curves') axes[2].legend() plt.tight_layout() return fig ``` ### Pattern 3: Time Series Analysis Visualization ```python def timeseries_dashboard(df, date_col, value_col): fig, axes = plt.subplots(3, 2, figsize=(14, 12)) # Raw time series axes[0, 0].plot(df[date_col], df[value_col], linewidth=0.5) axes[0, 0].set_title('Time Series') # Rolling mean rolling_mean = df[value_col].rolling(window=30).mean() axes[0, 1].plot(df[date_col], rolling_mean, color='red', label='30-day MA') axes[0, 1].plot(df[date_col], df[value_col], alpha=0.3) axes[0, 1].set_title('30-Day Moving Average') axes[0, 1].legend() # Seasonal decomposition from statsmodels.tsa.seasonal import seasonal_decompose decomposition = seasonal_decompose(df[value_col], model='additive', period=365) axes[1, 0].plot(decomposition.trend) axes[1, 0].set_title('Trend') axes[1, 1].plot(decomposition.seasonal) axes[1, 1].set_title('Seasonal') # Distribution by period df['month'] = df[date_col].dt.month sns.boxplot(x='month', y=value_col, data=df, ax=axes[2, 0]) axes[2, 0].set_title('Monthly Distribution') # Autocorrelation from pandas.plotting import autocorrelation_plot autocorrelation_plot(df[value_col], ax=axes[2, 1]) plt.tight_layout() return fig ```