correlation-explorer

You can calculate a correlation matrix across all numeric columns and generate an intuitive heatmap visualization. This Skill computes pairwise correlations using Pearson, Spearman, or Kendall coefficients and saves the visual output as a PNG file.

You can check statistical significance by computing optional p-values alongside your correlation coefficients. This allows you to assess whether the observed relationships between variables in your dataset are statistically significant.

Yes, you can use Spearman and Kendall correlation coefficients instead of Pearson. The Skill supports all three statistical methods, allowing you to choose the most appropriate coefficient for your data distribution.

The best way to detect multicollinearity is to compute a correlation matrix across your features and review the heatmap. This reveals strongly correlated variables that could destabilize your predictive model, facilitating feature selection.

Yes, you need pandas and scipy to compute p-values, along with matplotlib and seaborn to generate heatmaps. These dependencies provide the statistical and visualization functions required to calculate correlations and render visual outputs.