[Python] Why does the divergence value at the left and right boundaries show up as 1 instead...

I have tried two different ways of calculating a vector's divergence using Fipy: one using the gradient of its components and the other using face divergence. For simplicity, I have used a vector such that it has a constant x-component and a y-component is zero. In either case, the value of divergence at the left and right edge is different than the expected value. Please, help me understand and fix this. Below is the code for the check

import matplotlib.pyplot as plt
from fipy import Grid2D, CellVariable, FaceVariable

# Define the mesh
nx, ny = 100, 100 # Number of cells
dx, dy = 0.1, 0.1 # Cell size
mesh = Grid2D(nx=nx, ny=ny, dx=dx, dy=dy)

# Get cell center coordinates
x, y = mesh.cellCenters

# Define constant field variables
C = 2 # Constant gradient of ux along x axis
D = 0 # Constant uy
ux = CellVariable(name="ux", mesh=mesh, value=C*x)
uy = CellVariable(name="uy", mesh=mesh, value=D)

# Define the vector field
uvec = (ux, uy)

# Compute divergence using gradient components
div_u = CellVariable(name="div(u)", mesh=mesh, value=(ux.grad[0] + uy.grad[1]))

# Compute divergence using face variable
ux_face = ux.arithmeticFaceValue
uy_face = uy.arithmeticFaceValue
u = FaceVariable(mesh=mesh, value=(ux_face, uy_face))
facediv_u = CellVariable(name="face_div(u)", mesh=mesh, value=u.divergence)

print(div_u, "gradcomponents")
print(facediv_u, "facedivergence")


which gives the following result;

[1. 2. 2. ... 2. 2. 1.] gradcomponents
[1. 2. 2. ... 2. 2. 1.] facedivergence

Continue reading...