A simple to "path" ElasticNet parameters ( in python)

import numpy as np
import sys
import math

trfile=sys.argv[1]
tlabel=sys.argv[2]
parafile=sys.argv[3]


from sklearn.linear_model import ElasticNetCV

A=np.genfromtxt(trfile);
label=np.loadtxt(tlabel,dtype=np.int8);

clscv=ElasticNetCV(cv=3, l1_ratio=0.9);
clscv.fit(A,label)

y="%s %s" %(clscv.alpha_ ,clscv.l1_ratio)
f = file(parafile, 'w') # open for 'w'riting
f.write(y);
f.close


print y

trfile : training data

tlabel:label data

parafile: file to save parameters

two parameters: alpha and l1_ratio

but............set l1_ratio = 0.9

A better method will update to combine l1_ratio inside :)