Troubling to convert a 3d plot from Matlab to Python!

Hi everyone,

I have a problem in transforming a 3D plot from Matlab to Python. Here are my Matlab and Python codes. The output in Python is "Argument Z must be 2-dimensional." I still do not understand the framework in Matlab working in creating such a plot like this. Nothing that a T matrix has the shape (15, 25, 650). Are there differences between mesh in Matlab and np.meshgrid in Python? I would love it if there are ideas to help me solve this problem. Thank you!

#Matlab Code


m=25; n=15;

x(1)=0; 

for i=2:m+1
    
    dx=200; 

    x(i)=x(i-1)+dx;

end

y(1)=0;

for j=2:n+1

    dy=100;

    y(j)=y(j-1)+dy;

end

%Central mesh position

for j=1:n

    h(j)=(y(j)+y(j+1))/2;

end

for i=1:m

    z(i)=(x(i)+x(i+1))/2;

end

% Boundary condition

for j=1:n

    for i=1:m

        T(j,i,1)=0;

    end

end

mesh(z,h,T(:,:,1));

Python code

m = 25
n = 15
tmax = 650

x = np.zeros(m);
y = np.zeros(n);
h = np.zeros(n);
z = np.zeros(m);
T = np.zeros((n,m,tmax))

x[0] = 0
for i in range(1, m):
    x[i] = x[i-1] + 200

y[0] = 0
for j in range(1, n):
    y[j] = y[j-1] + 100


for j in range(1, n):
    h[j] = (y[j-1] + y[j])/2
    
for i in range(1, m):
    z[i] = (x[i] + x[i-1])/2

for j in range(1, n):
    for i in range(1, m):
        T[j, i, 1] = 0
        
X, Y, Z = np.meshgrid(z, h, T)
fig = plt.figure(figsize=(14,8))
ax = fig.gca(projection='3d')
surf = ax.plot_surface(X, Y, Z, cmap=cm.coolwarm,
                       linewidth=0, antialiased=False)
fig.colorbar(surf, shrink=0.5, aspect=5)

plt.show()

Answer 1

[surya venu]

Hi,

The error you're encountering arises because "np.meshgrid" in Python and "mesh" in MATLAB have some differences in how they handle dimensions and inputs. In your case, "Z" should indeed be a 2-dimensional array for the "plot_surface" function in Matplotlib.

Here's a corrected version of your Python code that should work:

import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm

m = 25
n = 15
tmax = 650
x = np.zeros(m)
y = np.zeros(n)
h = np.zeros(n)
z = np.zeros(m)
T = np.zeros((n, m, tmax))

x[0] = 0
for i in range(1, m):
    x[i] = x[i-1] + 200

y[0] = 0
for j in range(1, n):
    y[j] = y[j-1] + 100

for j in range(n):
    h[j] = (y[j] + y[j-1])/2 if j > 0 else y[j]/2

for i in range(m):
    z[i] = (x[i] + x[i-1])/2 if i > 0 else x[i]/2

for j in range(n):
    for i in range(m):
        T[j, i, 0] = 0  # Note: Use 0 instead of 1 for the first time step

# Create the meshgrid
X, Y = np.meshgrid(z, h)

# Extract the first time slice of T
Z = T[:, :, 0]

fig = plt.figure(figsize=(14, 8))
ax = fig.add_subplot(projection='3d')
surf = ax.plot_surface(X, Y, Z, cmap=cm.coolwarm, linewidth=0, antialiased=False)
fig.colorbar(surf, shrink=0.5, aspect=5)
plt.show()

Explanation of Differences:

MATLAB "mesh":

• In MATLAB, mesh(X, Y, Z) assumes X, Y, and Z are 2D matrices that define the grid and the surface height.

To know more about "mesh" function, check out:

https://www.mathworks.com/help/matlab/ref/mesh.html

Python "np.meshgrid":

• "np.meshgrid" in Python creates coordinate matrices from coordinate vectors. The resulting X and Y are 2D arrays suitable for surface plotting.

To know more about "meshgrid" function, check out:

https://numpy.org/doc/stable/reference/generated/numpy.meshgrid.html

Hope it helps