---

name: ds-observational-studies compatibility: opencode completeness: 95 content-types:

• code
• guidance
• do-dont
• examples description: '"Analyzes observational data using matching methods, propensity scores stratification, and adjustment for confounding bias"' license: MIT maturity: stable metadata: domain: coding output-format: code related-skills: ds-causal-inference, ds-instrumental-variables, ds-intervention-analysis ds-randomized-experiments ds-synthetic-control role: implementation scope: implementation triggers: observational studies, propensity score, matching, stratification, observational data, adjustment 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"

---

Observational Studies

Comprehensive guide to observational studies in machine learning and data science workflows.

When to Use This Skill

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

When NOT to Use This Skill

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

Observational Studies 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 Observational Studies

import pandas as pd
import numpy as np
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score, roc_auc_score

def compute_propensity_scores(df: pd.DataFrame, treatment_col: str, covariates: list) -> dict:
    """Compute propensity scores and evaluate model performance."""
    if treatment_col not in df.columns:
        raise ValueError(f"Treatment column '{treatment_col}' not found in DataFrame.")
    
    X = df[covariates].dropna()
    y = df.loc[X.index, treatment_col]
    
    model = LogisticRegression(max_iter=1000, random_state=42)
    model.fit(X, y)
    scores = pd.Series(model.predict_proba(X)[:, 1], index=X.index, name='propensity_score')
    
    y_pred = model.predict(X)
    metrics = {
        'accuracy': accuracy_score(y, y_pred)
        'roc_auc': roc_auc_score(y, scores)
    }
    return {'propensity_scores': scores, 'model_metrics': metrics}

# Example usage with synthetic data
if __name__ == "__main__":
    np.random.seed(42)
    n_samples = 500
    data = pd.DataFrame({
        'age': np.random.normal(50, 10, n_samples)
        'income': np.random.normal(60000, 15000, n_samples)
        'treatment': np.random.binomial(1, 0.5, n_samples)
    })
    covariates = ['age', 'income']
    result = compute_propensity_scores(data, 'treatment', covariates)
    print(f"Computed scores for {len(result['propensity_scores'])} samples")
    print(f"Model Metrics: {result['model_metrics']}")

Pattern 2: Production-Ready Observational Studies

import logging
from typing import Any, Dict, List
import pandas as pd
import numpy as np
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import NearestNeighbors

logger = logging.getLogger(__name__)

class ObservationalStudyPipeline:
    """Production implementation for observational study analysis."""
    
    def __init__(self, propensity_model=None, matching_method: str = 'nearest_neighbor'):
        self.propensity_model = propensity_model or LogisticRegression(max_iter=1000, random_state=42)
        self.matching_method = matching_method
        self.results: Dict[str, Any] = {}

    def _compute_propensity(self, df: pd.DataFrame, treatment_col: str, covariates: List[str]) -> pd.Series:
        X = df[covariates].dropna()
        y = df.loc[X.index, treatment_col]
        self.propensity_model.fit(X, y)
        return pd.Series(self.propensity_model.predict_proba(X)[:, 1], index=X.index, name='propensity_score')

    def _match_treatments(self, df: pd.DataFrame, caliper: float = 0.05) -> pd.DataFrame:
        treated = df[df['treatment'] == 1].dropna(subset=['propensity_score'])
        control = df[df['treatment'] == 0].dropna(subset=['propensity_score'])
        
        if len(treated) == 0 or len(control) == 0:
            raise ValueError("No treated or control observations found after dropping NaNs.")
            
        nn = NearestNeighbors(n_neighbors=1)
        nn.fit(control[['propensity_score']])
        distances, indices = nn.kneighbors(treated[['propensity_score']])
        
        mask = distances.flatten() < caliper
        matched_control = control.iloc[indices.flatten()[mask]]
        matched_treated = treated.iloc[mask]
        
        return pd.concat([matched_treated, matched_control], ignore_index=True)

    def execute(self, data: pd.DataFrame, treatment_col: str, covariates: List[str]) -> Dict[str, Any]:
        logger.info("Starting observational study pipeline")
        if data.empty:
            raise ValueError("Input data cannot be empty")
            
        df = data.copy()
        df['propensity_score'] = self._compute_propensity(df, treatment_col, covariates)
        matched_df = self._match_treatments(df)
        
        self.results = {
            'status': 'success'
            'matched_samples': len(matched_df)
            'treated_count': len(matched_df[matched_df['treatment'] == 1])
            'control_count': len(matched_df[matched_df['treatment'] == 0])
            'matched_data': matched_df
        }
        logger.info(f"Pipeline completed. Matched {self.results['matched_samples']} samples.")
        return self.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.

• Observational Study — Wikipedia
• Propensity Score Matching (NIST)
• Causal Inference with Observational Data (Harvard Biostats)
• Matching Methods — Stanford Statistics
• Regression Discontinuity Design (MIT OpenCourseWare)