Python极坐标等高线图绘制求助：非均匀数据插值与报错问题

问题背景 #

现有20000组非均匀分布的XYZ实测数据（含极坐标角度分量），绘制极坐标等高线图时遇到以下问题：

• 尝试pet.contour失败
• 使用tricontourf得到的图效果差，需要更好的插值效果
• 使用scipy.interpolate.griddata时触发TypeError：griddata() got multiple values for argument 'method'，调整contourf/contour等方法仍报错，怀疑数组格式问题但不知如何修正

尝试的tricontourf代码 #

# 注：原代码导入pandas别名错误，此处修正为pd
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

df = pd.read_excel('......xlsx', ...)
data_arr = df.to_numpy()
direction = np.radians(data_arr[:,1])
values = data_arr[:,0]
zeniths = data_arr[:,2]
r, theta = np.meshgrid(zeniths, direction)
fig, ax = plt.subplots(subplot_kw=dict(projection='polar'))    
# 颜色刻度范围
lmin = min(data_arr[:,0])
lmax = max(data_arr[:,0])
colormap = plt.get_cmap('rainbow')
norm = plt.colors.Normalize(lmin, lmax)
ax.set_theta_zero_location("N")
ax.set_theta_direction(-1)
ax.tricontourf(direction, zeniths, values, 30, cmap=colormap)
plt.show()

报错的griddata代码及错误信息 #

代码 #

# 注：原代码导入pandas别名错误，此处修正为pd
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from scipy import interpolate

df = pd.read_excel('...xlsx')
data_arr = df.to_numpy()
direction = np.radians(data_arr[:,1])
values = data_arr[:,0]
zeniths = data_arr[:,2]

def grid(direction, zeniths, values, resX=100, resY=100):
    xi = np.linspace(min(direction), max(direction), resX)
    yi = np.linspace(min(zeniths), max(zeniths), resY)
    Z = interpolate.griddata(direction, zeniths, values, xi, yi, method='cubic')
    X, Y = np.meshgrid(xi, yi)
    return X, Y, Z

X, Y, Z = grid(direction, zeniths, values)
contour = plt.contourf(X,Y,Z)

错误信息 #

Exception has occurred: TypeError
griddata() got multiple values for argument 'method'
File “…Contour polarPlot 2", line 20, in 
Z = interpolate.griddata(direction, zeniths, values, xi, yi, method='cubic')
   
File "/ Contour polarPlot 2", line 23, in <module>
X, Y, Z = grid(direction, zeniths, values)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: griddata() got multiple values for argument 'method'

问题解决 #

1. griddata参数错误修正 #

scipy.interpolate.griddata的正确参数格式要求：

• 第一个参数是**(N,2)的二维数组**，组合所有原始点的坐标
• 插值网格需以(Xi, Yi)形式传入（meshgrid生成的网格对）

修正后的grid函数：

def grid(direction, zeniths, values, resX=100, resY=100):
    # 组合原始点坐标为(N,2)数组
    points = np.column_stack((direction, zeniths))
    # 生成插值网格
    xi = np.linspace(direction.min(), direction.max(), resX)
    yi = np.linspace(zeniths.min(), zeniths.max(), resY)
    Xi, Yi = np.meshgrid(xi, yi)
    # 执行插值
    Z = interpolate.griddata(points, values, (Xi, Yi), method='cubic')
    return Xi, Yi, Z

2. 极坐标等高线图绘制优化 #

修正插值后，结合极坐标投影绘制高质量等高线图：

# 调用修正后的grid函数
theta_grid, r_grid, Z_grid = grid(direction, zeniths, values)

fig, ax = plt.subplots(subplot_kw=dict(projection='polar'))
# 设置极坐标方向（北为0度，顺时针旋转）
ax.set_theta_zero_location("N")
ax.set_theta_direction(-1)
# 绘制等高线图
contour = ax.contourf(theta_grid, r_grid, Z_grid, 30, cmap='rainbow')
# 添加颜色条
plt.colorbar(contour)
plt.show()

3. 其他注意事项 #

• 导入pandas时不要用df作为别名，避免覆盖后续的DataFrame变量
• 若cubic插值效果仍不理想，可尝试linear或nearest方法，或提高插值分辨率（resX/resY）
• 确保角度已转换为弧度（原代码已处理，无需修改）

内容的提问来源于stack exchange，提问作者Slawo Lomnicki