vasco - Unveiling connections in your Postgres data landscape.

vasco

5 views

Unveiling connections in your Postgres data landscape.

Pitch

Vasco is a powerful Postgres extension that empowers data analysts to uncover and explore hidden correlations within their datasets. Utilizing advanced techniques such as the Maximal Information Coefficient (MIC), Vasco goes beyond traditional linear associations, enabling users to detect intricate relationships among variables. Experience a deeper understanding of your data today.

Description

vasco is an advanced PostgreSQL extension designed to uncover hidden patterns and correlations within your data, enabling you to leverage the power of the Maximal Information Coefficient (MIC) and the MINE family of statistics. With vasco, you can easily analyze relationships between variables, allowing for deeper insights and enhancing your data analytics capabilities.

Key Features

Uncover Hidden Patterns: Quickly discover strong associations within your data using the mic(X, Y) function, which detects non-linear relationships beyond traditional linear correlation analysis.
Advanced Correlation Analysis: Use the Maximal Information Coefficient (MIC) to measure the strength of association between pairs of variables, helping you identify significant connections within your datasets. For instance:
```
SELECT mic(X, Y);  -- identifies associations between columns
SELECT corr(X,Y);  -- linear correlation (standard Postgres)
```
Correlation Matrices: Generate comprehensive correlation matrices to assess relationships across multiple variables at once, helping to visualize and analyze interdependencies effectively:
```
SELECT vasco_corr_matrix('v_faang', 'mic_v_faang');
```
Extensive Metric Support: In addition to MIC, vasco supports a variety of metrics for detailed analysis, such as Maximum Asymmetry Score (MAS), Maximum Edge Value (MEV), and Total Information Coefficient (TIC). Each metric can provide unique insights into variable relationships:

Metric SQL Function Interpretation
MAS SELECT mas(X, Y) Measures relationship deviation from monotonicity
TIC SELECT tic(X, Y) Total information captured by the variables
Custom Estimator Support: Select from different algorithms for estimating MIC, including ApproxMIC and MIC_e, allowing flexibility based on your data analysis needs:
```
SET vasco.mic_estimator = ApproxMIC;
SET vasco.mic_estimator = MIC_e;
```
pgvector Compatibility: Utilize vasco with pgvector for enhanced capabilities when working with vector types in your analyses:
```
SELECT mic(ARRAY[0, 1.3, 2, ...]::float4[]::vector, ARRAY[0, 1.3, 2, ...]::float4[]::vector);
```

Metric	SQL Function	Interpretation
MAS	`SELECT mas(X, Y)`	Measures relationship deviation from monotonicity
TIC	`SELECT tic(X, Y)`	Total information captured by the variables

Explore with Ease

Gain insights into complex datasets with user-friendly SQL commands to further enhance your exploration of data:

Leverage vasco to analyze stock price data, compute correlations, and visualize relationships, all while utilizing comprehensive metrics to deepen your understanding.

Conclusion

Whether you are a data analyst, scientist, or researcher, vasco equips you with the tools to discover meaningful insights from your PostgreSQL database. Join the growing community of users who are transforming their data analytics strategies with vasco.