Ask HN: Exploring Relationships in a 200-variable journal: Seeking Advice

Hi , I’m working with my own journals dataset containing 200 variables, mostly consisting of count or binary values. Zero counts and 0 values are implied.

I’m using Naïve Bayes to categorise the data against mental, physical, and social well-being scores alongside ANOVA and scatterplots.

I’m curious about finding relationships within the 200 variables beyond the well-being data. So far, I’ve created a heatmap based on time-based correlations and identified around 900 pairs with linear correlations using point-biserial correlation.

Any suggestions on additional analyses or techniques I could explore?

Cheers.