python - 如何修复 "ValueError: Argument Z must be 2-dimensional"错误

Question

我正在尝试绘制通过一组 (X,Y,Z) 3D 点的 3D 表面，但出现 raise ValueError("Argument Z must be 2-dimensional.")

ax = Axes3D(fig)
X = df.index.values
Y = np.where(df['trans_count'].values>0,np.ma.log(df['trans_count'].values), 0)

X, Y = np.meshgrid(X, Y)

Z = np.where(df['gpu_trans_count'].values>0, np.log(df['gpu_trans_count'].values), 0)

print(X.shape) #(37,37)
print(Y.shape) #(37,37)
print(Z.shape) #(37,)

ax.plot_surface(X, Y, Z)

ValueError:参数 Z 必须是二维的

Answer 1

您可以使用 np.expand_dims() 扩展 Z 的维度。您尚未指定要展开的轴。因此，我已经为两者提供了代码片段。只需根据您的意愿取消注释即可。

ax = Axes3D(fig)
X = df.index.values


Y = np.where(df['trans_count'].values>0,np.ma.log(df['trans_count'].values), 0)

X, Y = np.meshgrid(X, Y)


Z = np.where(df['gpu_trans_count'].values>0, np.log(df['gpu_trans_count'].values), 0)

#If you want to expand the x dimensions
#From (n,) to (1,n)
Z=np.expand_dims(Z,axis=0)
print(Z.shape) #(1,37)

# If you want to expand the y dim
# from (n,) to (n,1)
#Z=np.expand_dims(Z,axis=1)
#print(Z.shape) #(37,1)

print(X.shape) #(37,37)
print(Y.shape) #(37,37)



ax.plot_surface(X, Y, Z)