*cor*function of the

*stat*package to calculate correlation coefficient between variables. Second, he can use functions such as

*pairs*

*(graphics)*to visually check possible correlated variables. Third, he can combine the first two approach following the example of vinux in stackoverflow

*or using*

*ggpairs*function of

*GGally*package.

*First Approach*

Sepal.Length Sepal.Width Petal.Length Petal.Width

Sepal.Length 1.0000000 -0.1175698 0.8717538 0.8179411

Sepal.Width -0.1175698 1.0000000 -0.4284401 -0.3661259

Petal.Length 0.8717538 -0.4284401 1.0000000 0.9628654

Petal.Width 0.8179411 -0.3661259 0.9628654 1.0000000

*Second Approach*

*Third Approach*
This comment has been removed by the author.

ReplyDeleteAnother way to look at correlation is with correlograms. An overview is here: http://www.statmethods.net/advgraphs/correlograms.html

ReplyDeleteTry

corrgram(iris, upper.panel=panel.pts, lower.panel=panel.ellipse, diag.panel=panel.density)

This comment has been removed by the author.

ReplyDeleteHello, you show us three great approaches for correlations, thanks! I wonder about two optional things.

ReplyDelete1) In third approach, is there a possible set up which marks all significant correlations with * / ** / ***, depending on the given significance niveau?

2) (General question) Does it make sense to add a regression line into each correlation diagram, and if yes (specific question), how can this be done best way (e.g. in solution 3)?