listD[i] = polyListOptim[i].getMeasure()
G, h = polyListOptim[207].getInequality()
G.dot(polyListOptim[207].getLocation()) - h.flatten()
polyObj = polyOperation.PolygonObj(optimTestFun.rosen, A, b)
polyList = polyOperation.divideGivenPolygon(optimTestFun.rosen, polyObj)
polyObj = polyListOptim[25]
polyObj._location
hull = scipy.spatial.ConvexHull(polyObj._V)
x, sol, G, h = polyOperation.findAnalyticCenter(hull.equations[:, 0:2],
-hull.equations[:, 2],
full_output=True)
for i in potentialIndex:
print polyListOptim[i].getLocation()
print polyListOptim[i].hasSplit()
print polyListOptim[6].getVertices()
print polyListOptim[6].getLocation()
for o in polyListOptim:
print o.hasSplit()
minIndex = directUtil.findLowestObjIndex(polyListOptim)
polyListOptim[minIndex].getFx()
# Gx \precceq h
# numpy.array(h).flatten() - numpy.array(G).dot(x)
# so we are expecting everything to be positive
print numpy.array(h).flatten() - numpy.array(G).dot(x)
redundantIndex = polyOperation.redundantConstraintBox(boxBounds, A, b)
bindingIndex = polyOperation.bindingConstraintBox(boxBounds, A, b)
origin = h - G * matrix(x)
bindingIndex, hull, newG, newh = polyOperation.bindingConstraintBox(boxBounds, A, b, full_output=True)
x1, sol1, G1, h1 = polyOperation.findAnalyticCenter(
newG[bindingIndex.tolist(), :], newh[bindingIndex.tolist()], full_output=True
)
# first we find the origins
origin = h - G * matrix(x1)
# construct the set of points
D = mul(origin[:, [0, 0]] ** -1, G)
import scipy.spatial
import matplotlib.pyplot as plt
hull = scipy.spatial.ConvexHull(D)
points = numpy.array(D)
# Gx \precceq h
# numpy.array(h).flatten() - numpy.array(G).dot(x)
# so we are expecting everything to be positive
print numpy.array(h).flatten() - numpy.array(G).dot(x)
redundantIndex = polyOperation.redundantConstraintBox(boxBounds, A, b)
bindingIndex = polyOperation.bindingConstraintBox(boxBounds, A, b)
origin = h - G * matrix(x)
bindingIndex, hull, newG, newh = polyOperation.bindingConstraintBox(
boxBounds, A, b, full_output=True)
x1, sol1, G1, h1 = polyOperation.findAnalyticCenter(
newG[bindingIndex.tolist(), :],
newh[bindingIndex.tolist()],
full_output=True)
# first we find the origins
origin = h - G * matrix(x1)
# construct the set of points
D = mul(origin[:, [0, 0]]**-1, G)
import scipy.spatial
import matplotlib.pyplot as plt
hull = scipy.spatial.ConvexHull(D)
points = numpy.array(D)
G,h = polyListOptim[207].getInequality()
G.dot(polyListOptim[207].getLocation()) - h.flatten()
polyObj = polyOperation.PolygonObj(optimTestFun.rosen,A,b)
polyList = polyOperation.divideGivenPolygon(optimTestFun.rosen,polyObj)
polyObj = polyListOptim[25]
polyObj._location
hull = scipy.spatial.ConvexHull(polyObj._V)
x,sol,G,h = polyOperation.findAnalyticCenter(hull.equations[:,0:2],-hull.equations[:,2],full_output=True)
for i in potentialIndex:
print polyListOptim[i].getLocation()
print polyListOptim[i].hasSplit()
print polyListOptim[6].getVertices()
print polyListOptim[6].getLocation()
for o in polyListOptim:
print o.hasSplit()
minIndex = directUtil.findLowestObjIndex(polyListOptim)