Cannot initialize SignomialInequality outside of a SignomialsEnabled environment

[data25]

Above is the snippet of my code of a Signomial problem. While executing the code, i get the error "Cannot initialize SignomialInequality outside of SignomialsEnabled environment". I am not sure if i am missing anything. Help is much appreciated.

Thanks

[1ozturkbe]

Can you either post the code above as actual text (see this )or try to create a minimum working example (MWE)? Nothing looks obviously wrong at the moment.

[data25]

Hi, here is the code.

import numpy as np
import math
from gpkit import Variable, Model, SignomialsEnabled

#Decision variables
C_D = Variable("C_D", "-", "Drag coefficient")
C_L = Variable("C_L", "-", "Lift coefficent")
a = Variable("a", "-", "coefficient a")
b = Variable("b", "-", "coefficient b")
c = Variable("c", "-", "coefficient c")
alpha = Variable("alpha", "-", "angle of attack")

objective = C_D/C_L #to maximize cl/cd, we minimize its reciprocal
#Enabling signomials for subtraction
with SignomialsEnabled():
    constraints = [1<=0.00929839*(C_L**4.13477/C_D**0.0213768)+0.904471*(C_L**0.00219711/C_D**0.0213768)+1.4992e-06*(C_L**50.9294/C_D**0.0213768), 1<=(math.pi/C_L)*(2*alpha-15*(a/4)+3*b-2*c), 1<=a*(15/0.00004)-3*(b/0.00001)+2*(c/0.00001), 1>=0.0000001/C_D, -1>=(20/alpha), 1<=(20/alpha), 1>=(0.00001/a), 1<=(0.080780/a), 1<=(0.147875/b), 1>=(0.00001/b), 1>=(0.00001/c), 1<=(0.067095/c)]

m = Model(objective, constraints)
sol = m.localsolve(verbosity=0) #uses 'cvxopt'
print(sol.summary())

[1ozturkbe]

You do realize that you have a -1 >= signomial constraint for alpha, right? This is not possible by definition of the GP, since signomials and variables are restricted to the positive orthant.

[1ozturkbe]

Once you fix that issue, the model closes no problem. The way to overcome the negative variable issue is to create an offset (alpha=1 is actually -6 etc) or add other dummy variables. I hope this helps.

[data25]

thanks! it did work. the -1 was a typo.