python - Numba returns NotImplemented Error - cell vars are not supported -

somehow numba not want compile function. have suspicion might connected lambda functions internet revealed nothing. error code is:

numba returns notimplemented error - cell vars not supported

my file:

from __future__ import division scipy import special math import exp,sqrt import numba nb import numpy np  @nb.jit def build_cov_matrix(n,t,h,rho):      # initialisation of variables      sigma   = np.zeros((2*n,2*n))           # initialise covariance matrix sigma     t_grid  = np.linspace(t/n,t,n)          # time grid     gamma   = 0.5-h                         # simplifying parameter in calculations     d_h     = sqrt(2*h)/(h+0.5)             # factor used in computation      # definition of auxiliary functions      itt = lambda x: t_grid[x%n]             # translate index time      g   = lambda x: 2*h*(1/(1-gamma)*x**(-gamma)+gamma/(1-gamma)*x**(-1-gamma)*                     special.hyp2f1(1,1+gamma,3-gamma,1/x)/(2-gamma))      # covariance matrix building without z_0 , w\tilde_0       i,j in np.ndindex(sigma.shape):         if i<n , j<n:             sigma[i,j]=min(itt(i),itt(j))         elif i>=n , j>=n:             if i==j:                 sigma[i,j]=itt(i)**(2*h)             else:                 max_time=max(itt(i),itt(j))                 min_time=min(itt(i),itt(j))                 sigma[i,j]=min_time**(2*h)*g(max_time/min_time)         elif i<n , j>=n:             sigma[i,j]=rho*d_h*(itt(j)**(h+1/2)-(itt(j)-min(itt(i),itt(j)))**(h+1/2))         elif i>=n , j<n:             sigma[i,j]=rho*d_h*(itt(i)**(h+1/2)-(itt(i)-min(itt(i),itt(j)))**(h+1/2))      return sigma 

*note: aware code has indented didn't know how simultaneously lines , lazy manually, sorry. *