def testobjecterror(value):
try:
pytave.feval(1, "test_return", value)
print "FAIL:", (value, )
except pytave.ObjectConvertError:
pass
except Exception, e:
print "FAIL", (value, ), e
def testvalueerror(*value):
try:
pytave.feval(1, *value)
fail(value)
except pytave.ValueConvertError:
pass
except Exception, e:
fail(value, e)
def example(name,N,rel_err,eta=1):
# select the problem
if name in [ 'phillips','shaw' ]:
(A,b,x) = pytave.feval(3,name,N)
elif name=='ilaplace':
(A,b,x,tLAP) = pytave.feval(4,'i_laplace',N,1)
else:
raise Exception("I don't know problem "+name)
# 1 dimension arrays are 2dim in octave. Let's flat them!
b = b.flatten()
x = x.flatten()
# colors for the plot
colors = {
'real' : 'blue',
'tsvd' : 'red',
'tik' : 'green',
'newtik' : 'black'
}
# calculate the error and disturb b with it
err = numpy.linalg.norm(b,2) * rel_err
e = numpy.random.randn(b.size)
e = err / numpy.linalg.norm(e,2) *e
tilde_b = b + e
# dictionary for the computed solutions
tilde_x = {}
errors = {}
# regularize the problem, 'real' is a dummy method for the real solution
for method in [ 'real', 'tsvd', 'tik', 'newtik' ]:
if method is not 'real':
tilde_x[method] = reg.solve(A,tilde_b,eta*err,method)
errors[method] = (numpy.linalg.norm(tilde_x[method] - x,2))/numpy.linalg.norm(x,2)
#print ( method + " " + str((numpy.linalg.norm(tilde_x[method] - x,2))/numpy.linalg.norm(x,2)) )
else:
tilde_x[method] = x
#pylab.plot(tilde_x[method],colors[method],label=method)
# plot the solutions
for method in [ 'real', 'tsvd', 'tik', 'newtik' ]:
pylab.plot(tilde_x[method],colors[method],label=method)
if name is 'shaw':
pylab.legend(loc='upper left')
else:
pylab.legend(loc='upper right')
pylab.title(name+" " + str( rel_err*100 ) + "%")
pylab.savefig(name+'_'+str(int(rel_err*1000))+'.png')
#pylab.show()
pylab.clf()
return errors
def initialize():
"""Create initial sampling points."""
AiLHS = pytave.feval(1, "initial_pts_Nd", ninit, amin, amax, N)
AiLHS = numpy.array(AiLHS)
Ai = numpy.zeros([ninit,3])
for i in range(0,ninit):
x = pytave.feval(1, "find_near_pt",
AiLHS[0, i,:], N, spc, amin, amax)
x = numpy.array(x)
Ai[i,:] = x[0][0]
return Ai
def testexpect(value, expected):
try:
nvalue, = pytave.feval(1, "test_return", value)
if not equals(value, nvalue):
fail("sent in %s, expecting %s, got %s", (value, expected, nvalue))
except TypeError, e:
fail(value, e)
def testequal(value):
try:
nvalue, = pytave.feval(1, "test_return", value)
if not equals(value, nvalue):
fail("as %s != %s" % (value, nvalue))
except TypeError, e:
fail(value, e)
def ptilde(self, Pp, Pg):
""" Calculate transition probabilities
Parameters
------------
Pp : ndarray, shape (n, k)
Conditional choice probabilities for provinces
Pg : ndarray, shape (n, 2 k)
Conditional choice probabilities for the government
Returns
---------
P : ndarray
Transition probability matrix
Notes
-----------
Takes conditional choice probabilities :math:`P` as an input and
returns the transition matrix :math:`\\tilde{P}`.
This is a wrapper for the matlab function **Ptilde**.
"""
return pytave.feval(1, "Ptilde", Pp, Pg, self.model())[0]
def phigprov(self, Pp, Pg, theta):
""" Calculate transition probabilities
Parameters
------------
Pp : ndarray, shape (n, k)
Conditional choice probabilities for provinces
Pg : ndarray, shape (n, 2 k)
Conditional choice probabilities for the government
theta : ndarray, shape (5, )
Parameters
Returns
---------
V : ndarray
Observable state values
Notes
-----------
Takes conditional choice probabilities :math:`P` and :math:`\theta`
as an input and returns values :math:`V^P`.
This is the mapping :math:`\Phi` in part (b) of the assignment.
This is a wrapper for the matlab function **Phigprov**.
"""
theta = sp.atleast_2d(theta)
return pytave.feval(1, "Phigprov", Pp, Pg, theta, self.model())[0]
def new_p(self, Pp, Pg, theta):
""" Calculate transition probabilities
Parameters
--------------
Pp : ndarray, shape (n, k)
Conditional choice probabilities for provinces
Pg : ndarray, shape (n, 2 k)
Conditional choice probabilities for the government
theta : ndarray, shape (5, )
Parameters
Returns
---------
Pp : ndarray, shape (n, k)
New conditional choice probabilities for provinces
Pg : ndarray, shape (n, 2 k)
New conditional choice probabilities for the government
Notes
-----------
Takes conditional choice probabilities :math:`P` and :math:`\theta`
as an input and returns new conditional choice values.
This is the mapping :math:`\Psi` in part (c) of the assignment.
This is a wrapper for the matlab function **NewP**.
"""
theta = sp.atleast_2d(theta)
return pytave.feval(2, "NewP", Pp, Pg, theta, self.model())
def interpolate(self,points):
derivs = numpy.row_stack( [numpy.zeros((1,self.d)), numpy.eye(self.d)] )
resp = pytave.feval(1,'my_funeval',points.T,derivs)[0]
val = resp[:,:,0]
dval = resp[:,:,1:]
val = val.T
dval = numpy.rollaxis(dval,0,3)
return [val,dval]
def interpolate(self, points):
derivs = numpy.row_stack([numpy.zeros((1, self.d)), numpy.eye(self.d)])
resp = pytave.feval(1, 'my_funeval', points.T, derivs)[0]
val = resp[:, :, 0]
dval = resp[:, :, 1:]
val = val.T
dval = numpy.rollaxis(dval, 0, 3)
return [val, dval]
def testmatrix(value):
try:
nvalue, = pytave.feval(1, "test_return", value)
class1, = pytave.feval(1, "class", value)
class1 = class1.tostring()
class2, = pytave.feval(1, "class", nvalue)
class2 = class2.tostring()
if not equals(value, nvalue):
fail("as %s != %s" % (value, nvalue))
if value.shape != nvalue.shape:
fail("Size check failed for: %s. Expected shape %s, got %s with shape %s" \
%(value, value.shape, nvalue, nvalue.shape))
if class1 != class2:
fail("Type check failed for: %s. Expected %s. Got %s." %
(value, class1, class2))
except TypeError, e:
fail("Execute failed: %s" % value, e)
def __init__(self,atype, smin, smax, orders):
self.smin = smin
self.smax = smax
self.bounds = numpy.row_stack([smin,smax])
self.d = len(smin)
self.orders = orders
self.__interp__ = pytave.feval(1, 'create_space', atype, orders, smin,smax) # we don't really need it here..
self.grid = self.__interp__[0]['points'][0,0].T
def search(Aall, Jall, curr_bestA, theta, upb,
lob, N, amin, amax, spc, delta):
"""Search step."""
# make sure that all points are unique
(Am, Jm) = pytave.feval(2, "dsmerge", Aall, Jall)
next_ptsall = []
next_pts = None
max_no_searches = 100
no_searches = 0
while next_pts == None and no_searches < max_no_searches:
next_ptsall, min_est, max_mse_pt = pytave.feval(
3, "krig_min_find_MADS_oct",
Am, Jm, curr_bestA, theta, upb, lob,
N, amin, amax, spc, delta)
next_pts = check_for_new_points(next_ptsall, Aall)
no_searches += 1
return next_pts
def poll(Aall, Jall, curr_bestA, N, delta, spc, amin, amax):
"""Perform the poll step."""
poll_pts = None
try:
poll_ptsall = pytave.feval(
1, "MADS_poll_ptsNd3_oct",
curr_bestA, N, delta, spc, amin, amax)
poll_ptsall = poll_ptsall[0]
poll_pts = check_for_new_points(poll_ptsall, Aall)
poll_pts = pytave.feval(
1, "sort_poll_pts", Aall, Jall, poll_pts, theta,
upb, lob, N, amin, amax)
except Exception as e:
print " This poll DID NOT work, must refine."
return poll_pts
def __init__(self, atype, smin, smax, orders):
self.smin = smin
self.smax = smax
self.bounds = numpy.row_stack([smin, smax])
self.d = len(smin)
self.orders = orders
self.__interp__ = pytave.feval(1, 'create_space', atype, orders, smin,
smax) # we don't really need it here..
self.grid = self.__interp__[0]['points'][0, 0].T
def testsetget(variables, name, value):
try:
variables[name] = value
if name not in variables:
fail("set/get: %s: Should exist, not there." % name)
result, = pytave.feval(1, "isequal", value, variables[name])
if not result:
fail("set/get: %s -> %s: results diverged" % (name, value))
except Exception, e:
fail("set/get: %s" % name, e)
def eigenvalue_problem(A, M):
"""Solve the eigenvalue problem with first available backend."""
if backends[0] == "oct2py":
import oct2py as op
octave = op.Oct2Py()
e, l = octave.eig(A.array(), M.array())
elif backends[0] == "pytave":
import pytave
e, l = pytave.feval(2, "eig", A.array(), M.array())
return e, l
def get_smoothed_fieldmap(fm, fmmag, epi_qform, epi_shape):
import pytave
# FIXME: how do I automatically get the correct directory here?
pytave.addpath('/home/bobd/git/nims/nimsutil')
xform = np.dot(np.linalg.inv(fm.get_qform()), epi_qform)
fm_pixdim = np.array(fm.get_header().get_zooms()[0:3])
fm_mag_data = fm_mag.get_data().mean(3).squeeze()
# clean up with a little median filter and some greyscale open/close.
# We'll use a structure element that is about 5mm (rounded up to the nearest pixdim)
filter_size = 5.0/fm_pixdim
fm_mag_data = ndimage.median_filter(fm_mag_data, filter_size.round().astype(int))
fm_mag_data = ndimage.morphology.grey_opening(fm_mag_data, filter_size.round().astype(int))
fm_mag_data = ndimage.morphology.grey_closing(fm_mag_data, filter_size.round().astype(int))
# Total image volume, in cc:
fm_volume = np.prod(fm_pixdim) * np.prod(fm.get_shape()) / 1000
# typical human cranial volume is up to 1800cc. There's also some scalp and maybe neck,
# so we'll say 2500cc of expected tissue volume.
mag_thresh = np.percentile(fm_mag_data, max(0.0,100.0*(1.0-2500.0/fm_volume)))
mask = ndimage.binary_opening(fm_mag_data>mag_thresh, iterations=2)
# Now delete all the small objects, just keeping the largest (which should be the brain!)
label_im,num_objects = ndimage.measurements.label(mask)
h,b = np.histogram(label_im,num_objects)
mask = label_im==b[h==max(h[1:-1])]
mask_volume = np.prod(fm_pixdim) * max(h[1:-1]) / 1000.0
mask_sm = ndimage.gaussian_filter(mask.astype(np.float), filter_size)
fm_Hz = fm.get_data().astype(np.float).squeeze()
fm_Hz_sm = ndimage.gaussian_filter(fm_Hz * mask_sm, filter_size/2)
fm_final = np.empty(epi_shape[0:3])
ndimage.affine_transform(fm_Hz_sm, xform[0:3,0:3], offset=xform[0:3,3], output_shape=epi_shape[0:3], output=fm_final)
[xxBc,yyBc,zzBc] = np.mgrid[0:epi_shape[0],0:epi_shape[1],0:epi_shape[2]] # grid the epi, get location of each sampled voxel
# Now apply the transform. The following is equivalent to dot(xform,coords), where "coords" is the
# list of all the (homogeneous) coords. Writing out the dot product like this is just faster and easier.
xxB = xxBc*xform[0,0] + yyBc*xform[0,1] + zzBc*xform[0,2] + xform[0,3]
yyB = xxBc*xform[1,0] + yyBc*xform[1,1] + zzBc*xform[1,2] + xform[1,3]
zzB = xxBc*xform[2,0] + yyBc*xform[2,1] + zzBc*xform[2,2] + xform[2,3]
# use local linear regression to smooth and interpolate the fieldmaps.
fm_smooth_param = 7.5/np.array(fm.get_header().get_zooms()[0:3]) # Want 7.5mm in voxels, so =7.5/mm_per_vox
fm_smooth = pytave.feval(1,'localregression3d',fm.get_data(),xxB+1,yyB+1,zzB+1,np.array([]),np.array([]),fm_smooth_param,fm_mag.get_data())[0]
return fm_smooth
def callMethod(self, msg):
curPath = sys.path[0]
logFilePath = curPath + "/contactjson.txt"
logging.basicConfig(filename=logFilePath, level=logging.DEBUG)
logging.debug("OLDLoading contact service")
outputPictureName = "picturem.png"
pytave.feval(0, "addpath", curPath)
pytave.feval(0, "addpath", curPath + "/kMeans")
pytave.feval(0, "addpath", curPath + "/pytave")
outputPicture = curPath + "/kMeansPictures/" + outputPictureName
pytave.feval(0, "exec", curPath + "/kMeansPictures/" + msg, outputPicture)
###TODO: implement compression algorithm here in order to put logging information during running
# A = pytave.feval(1, "imread", '/home/aprovodi/simpleserver/VLayouts/output/services/bird_small.png')
# B = A[0] / 255.0 # Divide by 255 so that all values are in the range 0 - 1
# os.mkdir("/home/aprovodi/simpleserver/VLayouts/temp/zhopa")
# pytave.feval(1, "imwrite", B, "/home/aprovodi/simpleserver/VLayouts/temp/picture.jpg")
# result = curPath + "/kMeansPictures/" + msg
return outputPictureName
def callMethod(self, msg):
curPath = sys.path[0]
logFilePath = curPath + "/contactjson.txt"
logging.basicConfig(filename=logFilePath, level=logging.DEBUG)
logging.debug("OLDLoading contact service")
outputPictureName = "picturem.png"
pytave.feval(0, "addpath", curPath)
pytave.feval(0, "addpath", curPath + "/kMeans")
pytave.feval(0, "addpath", curPath + "/pytave")
outputPicture = curPath + "/kMeansPictures/" + outputPictureName
pytave.feval(0, "exec", curPath + "/kMeansPictures/" + msg,
outputPicture)
###TODO: implement compression algorithm here in order to put logging information during running
# A = pytave.feval(1, "imread", '/home/aprovodi/simpleserver/VLayouts/output/services/bird_small.png')
# B = A[0] / 255.0 # Divide by 255 so that all values are in the range 0 - 1
# os.mkdir("/home/aprovodi/simpleserver/VLayouts/temp/zhopa")
# pytave.feval(1, "imwrite", B, "/home/aprovodi/simpleserver/VLayouts/temp/picture.jpg")
# result = curPath + "/kMeansPictures/" + msg
return outputPictureName
import numpy, pytave A = numpy.random.random([2, 2]) e = pytave.feval(1, "eig", A) print 'Eigenvalues:', e
if Ai.shape ==(3,):
curr_bestA = Ai
else:
curr_bestA = Ai[i,:]
curr = [curr_bestJ, curr_bestA, nit, neval, prev_it]
cost_improve = 1
improve = 1
if improve == 0 and nit > 1:
cost_improve = 0
elif nit == 1:
cost_improve = 1
return cost_improve, curr_bestA, curr_bestJ
# add SMF to the octave load path
pytave.feval(0, "addpath", "SMF")
# number of parameters
N = 3
# min and max values for the parameters
amin = numpy.array([100, 0.1, 0.1])
amax = numpy.array([10000, 1.0, 0.5 ])
# initial guess
curr_best = []
curr_bestJ = 1000000
curr_bestA = numpy.array([1000, 0.5, 0.5])
# resolution of the Latin hyper cube
spc = 5*numpy.ones([1,N])
arr2ch = Numeric.array(["abc", "def"], Numeric.Character)
arr1o = Numeric.array([[1.0, "abc", 2 + 3j]], Numeric.PyObject)
arr2o = Numeric.array([[1.0, "abc", 2 + 3j], [4.0, arr1i, "def"]],
Numeric.PyObject)
alimit_int32 = Numeric.array([[-2147483648, 2147483647]], Numeric.Int32)
alimit_int16 = Numeric.array([[-32768, 32767, -32769, 32768]], Numeric.Int16)
alimit_int8 = Numeric.array([[-128, 127, -129, 128]], Numeric.Int8)
alimit_uint8 = Numeric.array([[0, 255, -1, 256]], Numeric.UnsignedInt8)
# This eval call is not to be seen as a encouragement to use Pytave
# like this. Create a separate .m-file with your complex Octave code.
pytave.eval(0,
"function [result] = test_return(arg); result = arg; endfunction")
pytave.feval(1, "test_return", 1)
def equals(a, b):
return Numeric.alltrue(Numeric.ravel(a == b))
def fail(msg, exc=None):
print "FAIL:", msg
traceback.print_stack()
if exc is not None:
traceback.print_exc(exc)
print ""
def testequal(value):
bc = [None, None]
u, v = TrialFunction(V), TestFunction(V)
p, q = TrialFunction(Q), TestFunction(Q)
p00 = inner(u, v) * dx
a00 = div(u) * div(v) * dx + inner(u, v) * dx
a01 = div(v) * p * dx
a10 = div(u) * q * dx
a11 = p * q * dx
P00 = assemble(p00)
A00 = assemble(a00)
A10 = assemble(a10)
A01 = assemble(a01)
A11 = assemble(a11)
e = pytave.feval(1, "check_eigs_vs", A11.array(), A01.array(), A00.array(),
P00.array())
e = pytave.feval(1, "abs", e[0])
e = pytave.feval(1, "sort", e[0])
e = e[0]
print " condition number 1 - div RT ", e[-1] / e[0], e[-1] / e[1]
p00 = inner(u, v) * dx
a00 = inner(u, v) * dx
a01 = div(v) * p * dx
a10 = div(u) * q * dx
# a11 = p*q*dx + inner(grad(p),grad(q))*dx
alpha = 1.0
n = FacetNormal(mesh)
h = CellSize(mesh)
h_avg = (h('+') + h('-')) / 2
import numpy, pytave A = numpy.random.random([2, 2]) B = numpy.random.random([2, 2]) e, v = pytave.feval(2, "eig", A, B) print 'Eigenvalues:', e print 'Eigenvectors:', v
import numpy, pytave A = numpy.random.random([2,2]) e = pytave.feval(1, "eig", A) print 'Eigenvalues:', e
def testvalueok(*value):
try:
pytave.feval(1, *value)
except Exception, e:
print "FAIL", (value, ), e
def set_values(self,values):
pytave.feval(0,'set_coeffs',values.T)[0]
def __call__(self, points):
derivs = numpy.zeros((1,self.d))
values = pytave.feval(1,'my_funeval',points.T,derivs)[0]
return values.T
for N in Ns: mesh = UnitSquare(N, N) V = VectorFunctionSpace(mesh, "Lagrange", 2) Q = FunctionSpace(mesh, "Lagrange", 1) bc = [DirichletBC(V, Constant((0,0)), boundary), None] u, v = TrialFunction(V), TestFunction(V) p, q = TrialFunction(Q), TestFunction(Q) a00 = inner(epsilon(u), epsilon(v))*dx a01 = div(v)*p*dx a10 = div(u)*q*dx a11 = p*q*dx AA, AArhs = block_symmetric_assemble([[a00, a01], [a10, a11]], bcs=bc) e = pytave.feval(1, "check_eigs", AA[0,0].array(), AA[1,0].array(), AA[1,1].array()) e = pytave.feval(1, "abs", e[0]) e = pytave.feval(1, "sort", e[0]) e = e[0] print " condition number 1 - eps ", e[-1]/e[0], e[-1]/e[1] a00 = div(u)*div(v)*dx + inner(u,v)*dx a01 = div(v)*p*dx a10 = div(u)*q*dx a11 = p*q*dx AA, AArhs = block_symmetric_assemble([[a00, a01], [a10, a11]], bcs=bc)
def interpolate(self,points):
derivs = numpy.row_stack( [numpy.zeros((1,self.d)), numpy.eye(self.d)] )
resp = pytave.feval(1,'my_funeval',points.T,derivs)[0]
val = resp[:,:,0]
dval = resp[:,:,1:]
val = val.T
dval = numpy.rollaxis(dval,0,3)
return [val,dval]
if __name__ == '__main__':
pytave.feval(0,'addpath','/home/pablo/Programmation/compecon/CEtools')
pytave.feval(0,'addpath','/home/pablo/Documents/Research/Thesis/Chapter_1/code/')
orders = numpy.array( [ 5, 5 ] )
smin = numpy.array( [ 0.9, 0.5 ] )
smax = numpy.array( [ 1.1, 2 ] )
f = lambda x: x[0:1,:] * x[1:2,:]
interp = CEInterpolation('spli',smin,smax,orders)
values = f(interp.grid)
print values
for N in Ns:
mesh = UnitSquare(N, N)
V = VectorFunctionSpace(mesh, "Lagrange", 2)
Q = FunctionSpace(mesh, "Lagrange", 1)
bc = [DirichletBC(V, Constant((0, 0)), boundary), None]
u, v = TrialFunction(V), TestFunction(V)
p, q = TrialFunction(Q), TestFunction(Q)
a00 = inner(epsilon(u), epsilon(v)) * dx
a01 = div(v) * p * dx
a10 = div(u) * q * dx
a11 = p * q * dx
AA, AArhs = block_symmetric_assemble([[a00, a01], [a10, a11]], bcs=bc)
e = pytave.feval(1, "check_eigs", AA[0, 0].array(), AA[1, 0].array(),
AA[1, 1].array())
e = pytave.feval(1, "abs", e[0])
e = pytave.feval(1, "sort", e[0])
e = e[0]
print " condition number 1 - eps ", e[-1] / e[0], e[-1] / e[1]
a00 = div(u) * div(v) * dx + inner(u, v) * dx
a01 = div(v) * p * dx
a10 = div(u) * q * dx
a11 = p * q * dx
AA, AArhs = block_symmetric_assemble([[a00, a01], [a10, a11]], bcs=bc)
e = pytave.feval(1, "check_eigs", AA[0, 0].array(), AA[1, 0].array(),
AA[1, 1].array())
e = pytave.feval(1, "abs", e[0])
def fit_values(self, values):
pytave.feval(0, 'set_coeffs', values.T)[0]
p, q = TrialFunction(Q), TestFunction(Q)
p00 = inner(u,v)*dx
a00 = div(u)*div(v)*dx + inner(u,v)*dx
a01 = div(v)*p*dx
a10 = div(u)*q*dx
a11 = p*q*dx
P00 = assemble(p00)
A00 = assemble(a00)
A10 = assemble(a10)
A01 = assemble(a01)
A11 = assemble(a11)
e = pytave.feval(1, "check_eigs_vs", A11.array(), A01.array(), A00.array(), P00.array())
e = pytave.feval(1, "abs", e[0])
e = pytave.feval(1, "sort", e[0])
e = e[0]
print " condition number 1 - div RT ", e[-1]/e[0], e[-1]/e[1]
p00 = inner(u,v)*dx
a00 = inner(u,v)*dx
a01 = div(v)*p*dx
a10 = div(u)*q*dx
# a11 = p*q*dx + inner(grad(p),grad(q))*dx
alpha = 1.0
n = FacetNormal(mesh)
h = CellSize(mesh)
h_avg = (h('+') + h('-'))/2
#!/usr/bin/python
import pytave
try:
pytave.feval(
1,
"",
)
except pytave.OctaveError, e:
print "test ok"
except:
print "test fail"
try:
pytave.feval(
1,
"cell",
)
except pytave.ValueConvertError, e:
print "test ok"
except:
print "test fail"
try:
pytave.feval(1, "sin", {"asdf": "asdf"})
except pytave.ObjectConvertError, e:
print "test ok"
except:
print "test fail"
values = pytave.feval(1, 'my_funeval', points.T, derivs)[0]
return values.T
def interpolate(self, points):
derivs = numpy.row_stack([numpy.zeros((1, self.d)), numpy.eye(self.d)])
resp = pytave.feval(1, 'my_funeval', points.T, derivs)[0]
val = resp[:, :, 0]
dval = resp[:, :, 1:]
val = val.T
dval = numpy.rollaxis(dval, 0, 3)
return [val, dval]
if __name__ == '__main__':
pytave.feval(0, 'addpath', '/home/pablo/Programmation/compecon/CEtools')
pytave.feval(0, 'addpath',
'/home/pablo/Documents/Research/Thesis/Chapter_1/code/')
orders = numpy.array([5, 5])
smin = numpy.array([0.9, 0.5])
smax = numpy.array([1.1, 2])
f = lambda x: x[0:1, :] * x[1:2, :]
interp = CEInterpolation('spli', smin, smax, orders)
values = f(interp.grid)
print values
def __call__(self, points):
derivs = numpy.zeros((1, self.d))
values = pytave.feval(1, 'my_funeval', points.T, derivs)[0]
return values.T
import numpy, pytave A = numpy.random.random([2,2]) B = numpy.random.random([2,2]) e, v = pytave.feval(2, "eig", A, B) print 'Eigenvalues:', e print 'Eigenvectors:', v
