---

name: ds-confidence-intervals compatibility: opencode completeness: 95 content-types:

• code
• guidance
• do-dont
• examples description: '"Provides Constructs confidence intervals using bootstrap, analytical methods, and uncertainty quantification for parameter estimation"' license: MIT maturity: stable metadata: domain: coding output-format: code related-skills: ds-bayesian-inference, ds-hypothesis-testing, ds-monte-carlo role: implementation scope: implementation triggers: confidence intervals, bootstrap, uncertainty quantification, confidence bounds, credible intervals archetypes:
  • tactical
  • generation anti_triggers:
  • brainstorming
  • vague ideation
  • code golf
  • over-engineering response_profile: verbosity: low directive_strength: high abstraction_level: operational version: "1.0.0"

---

Confidence Intervals

Comprehensive guide to confidence intervals in machine learning and data science workflows.

When to Use This Skill

• Solving real-world statistical inference problems
• Building machine learning pipelines with confidence intervals
• Implementing best practices for confidence intervals
• Optimizing model performance using confidence intervals techniques
• Learning industry-standard approaches to confidence intervals

When NOT to Use This Skill

• When using pre-built libraries without understanding underlying concepts
• For toy problems that don't require confidence intervals 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

Confidence Intervals 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 Confidence Intervals

import numpy as np
import pandas as pd
from scipy import stats

def calculate_analytical_ci(data: np.ndarray, confidence: float = 0.95) -> tuple[float, float]:
    """Calculate analytical confidence interval using t-distribution."""
    if not isinstance(data, np.ndarray):
        data = np.asarray(data)
    n = len(data)
    if n < 2:
        raise ValueError("Sample size must be at least 2")
    mean = np.mean(data)
    std_err = stats.sem(data)
    margin = stats.t.ppf((1 + confidence) / 2, df=n - 1) * std_err
    return (mean - margin, mean + margin)

def calculate_bootstrap_ci(data: np.ndarray, confidence: float = 0.95, n_resamples: int = 1000) -> tuple[float, float]:
    """Calculate bootstrap confidence interval via resampling."""
    rng = np.random.default_rng(42)
    bootstrap_means = np.empty(n_resamples)
    for i in range(n_resamples):
        sample = rng.choice(data, size=len(data), replace=True)
        bootstrap_means[i] = np.mean(sample)
    alpha = 1 - confidence
    lower = np.percentile(bootstrap_means, 100 * alpha / 2)
    upper = np.percentile(bootstrap_means, 100 * (1 - alpha / 2))
    return (lower, upper)

Pattern 2: Production-Ready Confidence Intervals

import logging
import numpy as np
import pandas as pd
from typing import Any, Dict, Literal

logger = logging.getLogger(__name__)

class ConfidenceIntervalCalculator:
    """Production-grade calculator for confidence intervals."""
    
    def __init__(self, confidence: float = 0.95, method: Literal['analytical', 'bootstrap'] = 'analytical', n_resamples: int = 1000):
        self.confidence = confidence
        self.method = method
        self.n_resamples = n_resamples
        logger.info(f"Initialized CI calculator with method={method}, confidence={confidence}")

    def execute(self, data: pd.DataFrame, column: str) -> Dict[str, Any]:
        """Execute confidence interval calculation on specified column."""
        if column not in data.columns:
            raise KeyError(f"Column '{column}' not found in DataFrame")
        
        values = data[column].dropna().values
        if len(values) < 2:
            raise ValueError("Insufficient non-null data for CI calculation")
            
        try:
            if self.method == 'analytical':
                n = len(values)
                mean = np.mean(values)
                std_err = stats.sem(values)
                margin = stats.t.ppf((1 + self.confidence) / 2, df=n - 1) * std_err
                lower, upper = mean - margin, mean + margin
            else:
                rng = np.random.default_rng(42)
                boots = np.array([np.mean(rng.choice(values, size=len(values), replace=True)) 
                                  for _ in range(self.n_resamples)])
                alpha = 1 - self.confidence
                lower, upper = np.percentile(boots, 100 * alpha / 2), np.percentile(boots, 100 * (1 - alpha / 2))
                
            logger.info(f"Calculated CI: [{lower:.4f}, {upper:.4f}]")
            return {
                'status': 'success'
                'lower_bound': float(lower)
                'upper_bound': float(upper)
                'point_estimate': float(np.mean(values))
                'method': self.method
                'sample_size': int(n if self.method == 'analytical' else len(values))
            }
        except Exception as e:
            logger.error(f"CI calculation failed: {e}")
            return {'status': 'error', 'message': str(e)}

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
• ✅ Follow PEP 8 and SOLID principles for maintainable statistical code

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.

• Confidence Interval — Wikipedia
• SciPy Stats — Statistical Distributions
• Statistical Intervals (NIST Handbook)
• Confidence Interval Calculator (GraphPad)
• Bayesian Credible Intervals — PyMC docs