0

Can someone enlighten me ? I know it's very basic but i'm really confused with how to make correlation with numeric AND non-numeric variable.

I have a data.frame with multiple variables. Some of it are numerical and other non-numeric. My data.frame look something like this :

Var1  Var2  Var3   Var4   Var7(binary) 
0.1     1    one    0.9        1
3.3     3    one    0          0
4.4     4    two    2          0
5.5     5    one    3          1 
1.0     2    one    3.2        1

With this data.frame, I have multiple models I want to test. But, first, I want to check if certain variables are correlated. Can I just use Pearson test? I use pair.panel to check for correlated variable but I think it's using Pearson test and I'm not sure I can use Pearson for no numeric data ?

I'm very confused ...

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Moineau Maaka
  • 1
  • 1
    Could you explain what you mean by "correlated" (a term with a general meaning and several different specific statistical ones) and tell us how that would be related to your "multiple models"? What models would those be? Knowing them might help us interpret your meaning. – whuber Nov 06 '20 at 19:37
  • Does this answer your question? [Correlations with unordered categorical variables](https://stats.stackexchange.com/questions/108007/correlations-with-unordered-categorical-variables) – T.E.G. Nov 06 '20 at 19:37

0 Answers0