powered by
Derivative of MCP function as described in Fan and Li (2001).
mcp_deriv(x,lambda,a)
Derivative of MCP function with tuning parameters lambda and "a" evaluated at "x".
Number to be evaluated
Tuning parameter lambda
Tuning parameter a
Ben Sherwood
[1] Zhang, C. Nearly unbiased variable selection under minimax concave penalty. (2010). Ann. Statist., 38, 894--942.
mcp(3,1) mcp(-3,1) mcp(.001,2)
Run the code above in your browser using DataLab