Let's say I have some samples, that I gathered per second, like these

Number km Benzine Speed Weight

1 10 100 80 100
2 20 90 100 120
3 30 80 80 110

with thousand of samples and more than twenty features (like speed and weight).

I know that there is a correlation among km and benzine. As km increase, benzine decreases. I like to find if there is any relationship among the other features-factors and the km or the benzine or both of them. How can i find if there is a correlation and which is the strongest one?

I tried CCA and the Pyrcca but couldn't understand it. Is it possible? Is there any simple example to understand? Thank you