Ok, so I'm making a data set that correlates the beer consumption level of a country to the weather of the country, but the weather statistics I have don't correlate as much as I would like. These statistics are:

  • Maximum temperature of the day
  • Minimum temperature of the day
  • Mean temperature of the day
  • Precipitation

So, are there any new statistics I can calculate from these ones? (I would prefer actual weather statistics that are used by experts but if not don't worry, I'll accept all ideas :) )

P.D: Here's the ones I've already tried so these don't count:

  • MaxTemperature - MinTemperature
  • (MaxTemperature + MinTemperature)/2
  • $\begingroup$ With only those four, there's really nothing else. You only really have two types of variables, temperature and precip. Maybe you could combine precip and temp into some sort of "nice day" index, but we don't have anything official. Weather only somewhat correlates to most things, that's just how it goes :-/ $\endgroup$ Commented Oct 5, 2022 at 17:38
  • $\begingroup$ @gerard Asbert Marcos you can do a climatology time series analysis. Do some binning etc. Harmonic oscillations etc $\endgroup$
    – user1066
    Commented Oct 6, 2022 at 9:18

1 Answer 1


but the weather statistics I have don't correlate as much as I would like Blockquote

That sounds like you are trying to find evidence that suits your hypothesis, in stead of basing your conclusion on the data. If there is no correlation between the two, no matter which statistic you will throw at it will prove that there is one.

That being said, you can always get fancy with metrics like precipitation / temperature and all sorts of variations. I advice against this.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.