Summarize Strength of Relationships Using R-Squared From Linear Regression
Usage
r2describe(x, nvmax = 10)
Value
nothing
Arguments
x
numeric matrix with 2 or more columns
nvmax
maxmum number of columns of x to use in predicting a given column
Author
Frank Harrell
Details
Function to use leaps::regsubsets() to briefly describe which variables more strongly predict another variable. Variables are in a numeric matrix and are assumed to be transformed so that relationships are linear (e.g., using redun() or transcan().)