CODEWITHSUNDEEP
pythonpandasmatplotlib
Will
Will

Reputation: 651

matplotlib 3d surface displaying incorrect x and y data

Alright, I'm trying to plot a matplotlib 3d surface graph.

import pandas as pd
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import itertools as it
import numpy as np


plt.clf()
fig = plt.figure(figsize=(6.27, 9.69))
ax = Axes3D(fig)

x = list(range(1,101,11))
y = list(range(1,101,11))

df = pd.DataFrame(columns=y, index=x)
iterator = it.product(x,y)

for i in iterator:
    number = np.random.rand()
    df.loc[i[0],i[1]] = number

surf = ax.plot_surface(df.index,
                       df.columns,
                       df,
                       cmap='viridis_r',
                       linewidth=0.5,
                       alpha = 0.5,
                       edgecolor = 'k')
ax.set_xlim3d(0,101)
ax.set_ylim3d(0,101)
ax.invert_yaxis()

cb = fig.colorbar(surf, shrink=0.8 , alpha = 0.5)
cb.solids.set_edgecolor("face")
plt.show()

My code works (after a fashion) but all my results only appear on a straight line where x=y i.e. a 2d graph.

enter image description here

I had thought it might be something to do with my use of itertools (the code not cycling through all the iterations) but the dataframe seems fine.

Can anybody tell me what I'm doing wrong?

Upvotes: 0

Views: 319

Answers (1)

CsuGouv
CsuGouv

Reputation: 249

I checked, and you do need to compute a grid before plotting. You can just add something like :

...
for i in iterator:
    number = np.random.rand()
    df.loc[i[0],i[1]] = number

xv, yv = np.meshgrid(df.index, df.columns)

surf = ax.plot_surface(xv,
                       yv,
                       df,
                       cmap='viridis_r',
                       linewidth=0.5,
                       alpha = 0.5,
                       edgecolor = 'k')
...

I don't really know if this is the expected result, but you obtain a real surface ! meshgrid just creates matrices which correspond to different permutations of values from x and y, to allow you to have different x for one y and vice versa.

Upvotes: 1

Related Questions