Ds Data Visualization

Data Visualization

Comprehensive guide to data visualization in machine learning and data science workflows.

When to Use This Skill

• Solving real-world exploratory data analysis problems
• Building machine learning pipelines with data visualization
• Implementing best practices for data visualization
• Optimizing model performance using data visualization techniques
• Learning industry-standard approaches to data visualization

When NOT to Use This Skill

• When using pre-built libraries without understanding underlying concepts
• For toy problems that don't require data visualization rigor
• When domain expertise in specific problem requires different approach
• If your problem doesn't require the complexity this skill provides

Purpose and Key Concepts

Data Visualization is a critical component of the machine learning workflow. This skill covers:

1. Theoretical foundations — Mathematical principles and statistical concepts
2. Practical implementation — Working code examples and patterns
3. Common pitfalls — Mistakes to avoid and how to recover from them
4. Best practices — Industry-standard approaches and optimization techniques

Core Workflow

1. Understand the problem — Clearly define what you're solving for
2. Select approach — Choose the right technique for your data and constraints
3. Implement solution — Write clean, tested code following best practices
4. Validate results — Verify your implementation with tests and validation
5. Optimize performance — Improve efficiency and accuracy incrementally

Implementation Patterns

Pattern 1: Basic Data Visualization

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

def create_basic_visualizations(df: pd.DataFrame) -> dict:
    """Generate basic exploratory visualizations for a DataFrame."""
    if df.empty:
        raise ValueError("DataFrame cannot be empty")

    numeric_cols = df.select_dtypes(include=[np.number]).columns
    if len(numeric_cols) == 0:
        raise ValueError("No numeric columns found for visualization")

    fig, axes = plt.subplots(1, 2, figsize=(12, 5))

    # Histogram of first numeric column
    sns.histplot(df[numeric_cols[0]], kde=True, ax=axes[0], color='skyblue')
    axes[0].set_title(f'Distribution of {numeric_cols[0]}')

    # Boxplot for remaining numeric columns
    if len(numeric_cols) > 1:
        df[numeric_cols].plot.box(ax=axes[1])
        axes[1].set_title('Boxplots of Numeric Features')
    else:
        axes[1].text(0.5, 0.5, 'Need >1 numeric column for boxplot', ha='center')

    plt.tight_layout()
    return {'figure': fig, 'numeric_columns': numeric_cols.tolist()}

if __name__ == "__main__":
    sample_df = pd.DataFrame({
        'age': np.random.normal(35, 10, 200)
        'income': np.random.exponential(50000, 200)
        'score': np.random.randint(0, 100, 200)
    })
    result = create_basic_visualizations(sample_df)
    plt.show()
    print(f"Generated plots for columns: {result['numeric_columns']}")

Pattern 2: Production-Ready Data Visualization

import logging
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from typing import Any, Dict, List, Optional
import os

logger = logging.getLogger(__name__)

class DataVisualization:
    """Production implementation of Data Visualization"""
    
    def __init__(self, output_dir: Optional[str] = None, dpi: int = 150):
        self.output_dir = output_dir or "viz_output"
        self.dpi = dpi
        os.makedirs(self.output_dir, exist_ok=True)
        logger.info(f"Initialized DataVisualization with output_dir={self.output_dir}")
    
    def _validate_data(self, data: pd.DataFrame) -> None:
        if not isinstance(data, pd.DataFrame):
            raise TypeError("Input must be a pandas DataFrame")
        if data.empty:
            raise ValueError("Input DataFrame cannot be empty")
        numeric_cols = data.select_dtypes(include=[np.number]).columns
        if len(numeric_cols) == 0:
            raise ValueError("DataFrame must contain at least one numeric column")

    def execute(self, data: pd.DataFrame, target_col: Optional[str] = None) -> Dict[str, Any]:
        """Execute Data Visualization on data"""
        self._validate_data(data)
        results = {'plots_generated': [], 'status': 'success'}
        
        try:
            # 1. Correlation Heatmap
            fig_corr, ax_corr = plt.subplots(figsize=(8, 6))
            corr_matrix = data.corr(numeric_only=True)
            sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', ax=ax_corr, fmt='.2f')
            ax_corr.set_title('Feature Correlation Matrix')
            plt.tight_layout()
            corr_path = os.path.join(self.output_dir, 'correlation_heatmap.png')
            fig_corr.savefig(corr_path, dpi=self.dpi)
            plt.close(fig_corr)
            results['plots_generated'].append(corr_path)
            logger.info(f"Saved correlation heatmap to {corr_path}")

            # 2. Distribution Plots
            numeric_cols = data.select_dtypes(include=[np.number]).columns.tolist()
            for col in numeric_cols[:5]:
                fig_dist, ax_dist = plt.subplots(figsize=(6, 4))
                sns.histplot(data[col], kde=True, ax=ax_dist, color='teal')
                ax_dist.set_title(f'Distribution of {col}')
                plt.tight_layout()
                dist_path = os.path.join(self.output_dir, f'dist_{col}.png')
                fig_dist.savefig(dist_path, dpi=self.dpi)
                plt.close(fig_dist)
                results['plots_generated'].append(dist_path)
                
        except Exception as e:
            logger.error(f"Visualization pipeline failed: {str(e)}")
            results['status'] = 'failed'
            results['error'] = str(e)
            
        return results

Best Practices

• ✅ Always validate your implementation on test data
• ✅ Document your assumptions and methodology
• ✅ Use version control for reproducibility
• ✅ Monitor performance metrics in production
• ✅ Periodically review and update your approach
• ✅ Test with edge cases and outliers
• ✅ Log all significant operations for debugging

Common Pitfalls

| Pitfall | Problem | Solution | |

Constraints

MUST DO

• Validate all data preprocessing steps are fit-only on training data, never on validation or test sets
• Implement reproducible pipelines with fixed random seeds and deterministic operations where possible
• Report model performance with confidence intervals via bootstrapping or cross-validation across multiple runs
• Log all experiments with parameters, metrics, and artifacts using MLflow or equivalent tracking system

MUST NOT DO

• Do not evaluate a model on the same data used for training — always hold out a proper test set
• Avoid overfitting to the validation set by limiting hyperparameter search iterations
• Never use features that can only be computed at inference time (look-ahead bias)
• Do not report single-run accuracy without statistical significance testing or error bars

Live References

Authoritative documentation links for this skill's domain. The model follows markdown links at load time to resolve external references and inline content.

• Matplotlib Gallery
• Seaborn Documentation
• Plotly Python Documentation
• Bokeh Visualization Library
• Data Visualization Best Practices (Nightingale)