Hello everyone
With tabular data, could it make sense to use ‘year’ and/or ‘time’ as a continuous rather than a categorical type in order to extract any type of relationship from these variables?
I can’t add years, but they have an implicit order relationship.
Thanks
I think it’s worth trying, but I think years as categorical is more flexible than years as continuous (if you have reasonable number of different values). Continuous variables seem more appropriate then depended variable is more smooth between years (in this example). If you have data that can vary a lot between years it’s seems more reasonable to use categorical. For ex. you predict stock market value. It this case crisis year seem to have pretty different profile as year before (and year after as well with all these government money).
But it’s a case dependent matter