Data Analysis

Perform exploratory data analysis on [dataset]:

Dataset: [description]
Size: [rows/records]
Key variables: [list columns]
Business question: [what you're trying to understand]
Known issues: [data quality concerns]

EDA steps:
- Data structure and types
- Summary statistics (mean, median, std dev, etc.)
- Missing data analysis
- Distribution of key variables
- Outlier detection
- Correlation analysis
- Pattern identification
- Segmentation opportunities
- Data quality issues
- Visualization recommendations
- Initial insights
- Recommended next steps

Provide Python/R code suggestions.

Design a statistical test for [hypothesis]:

Hypothesis: [what you want to test]
Data: [available data]
Sample size: [n]
Variables: [dependent and independent]
Confidence level: [typically 95%]
Business context: [what decision this informs]

Test design:
- Null and alternative hypotheses
- Appropriate statistical test (t-test, ANOVA, chi-square, etc.)
- Assumptions to check
- Sample size adequacy
- Significance level (α)
- Power analysis
- Test procedure steps
- Interpretation guidelines
- Limitations and caveats
- Action thresholds
- Code implementation
- Reporting format

Create a data cleaning strategy for [dataset]:

Dataset: [description]
Size: [records and fields]
Source: [where data comes from]
Issues observed: [problems found]
Use case: [how data will be used]
Quality requirements: [standards needed]

Cleaning strategy:
- Data profiling results
- Missing value treatment (impute, delete, flag)
- Outlier handling approach
- Duplicate record resolution
- Data type corrections
- Standardization rules
- Validation rules
- Transformation needs
- Documentation of changes
- Quality metrics
- Before/after comparison
- Implementation code
- Automation opportunities

Design an anomaly detection system for [use case]:

Use case: [fraud, defects, outliers, etc.]
Data: [description of data]
Volume: [how much data]
Expected anomaly rate: [%]
Consequence of missing: [impact]
False positive tolerance: [acceptable rate]

Detection approach:
- Anomaly definition
- Detection method (statistical, ML, rule-based)
- Feature selection
- Threshold determination
- Model choice rationale
- Training approach
- Performance metrics
- Alert mechanism
- Validation strategy
- False positive handling
- Continuous learning
- Implementation plan
- Monitoring and maintenance

Analyze correlations in [dataset] for [purpose]:

Dataset: [description]
Variables of interest: [list]
Purpose: [prediction, explanation, discovery]
Sample size: [n]
Domain: [business context]

Analysis:
- Correlation matrix
- Strong correlations identified (>0.7 or <-0.7)
- Scatter plot recommendations
- Multicollinearity concerns
- Causation vs correlation clarification
- Confounding variables
- Lagged correlations (time series)
- Non-linear relationships
- Business interpretation
- Actionable insights
- Further investigation needed
- Visualization approach

Develop a customer segmentation strategy:

Customers: [size of customer base]
Data available: [behavioral, demographic, transactional]
Objective: [personalization, targeting, retention]
Current approach: [existing segments if any]
Constraints: [limitations]

Segmentation approach:
- Segmentation variables selection
- Methodology (RFM, K-means, hierarchical, etc.)
- Optimal number of segments
- Segment profiles and characteristics
- Segment sizes
- Segment naming/personas
- Differentiating behaviors
- Value by segment
- Actionability of segments
- Segment migration analysis
- Activation strategy
- Tracking and measurement
- Refresh frequency

Perform trend analysis on [metric/data]:

Metric: [what you're analyzing]
Time period: [duration and granularity]
Data points: [how many observations]
Seasonality: [expected patterns]
External factors: [known influences]
Decision: [what you'll do with insights]

Trend analysis:
- Time series visualization
- Trend direction and magnitude
- Trend decomposition (trend, seasonal, irregular)
- Change points identification
- Growth rate calculation
- Forecasting approach
- Confidence intervals
- Comparison to benchmarks
- Explanatory factors
- Leading indicators
- Scenario projections
- Recommendations
- Monitoring plan

Generate insights from [analysis/data]:

Analysis: [describe analysis performed]
Data: [summary of data]
Audience: [who will receive insights]
Decision context: [what's being decided]
Priorities: [business priorities]

Insight framework:
- Key findings (5-7 insights)
- So what? (business implication)
- Evidence supporting each insight
- Confidence level
- Surprising vs expected findings
- Actionable recommendations
- Quick wins identified
- Risks or caveats
- Data limitations
- Further analysis needed
- Storytelling structure
- Visualization recommendations

Format as executive summary.