我试图绘制此轮廓，直到#upperlimit坐标变为黑色，结果应该类似于图b（附加）figure
图A是下一个代码的输出：
from scipy.interpolate import griddata
import numpy as np
import matplotlib.pyplot as plt
M = np.array([[0.000000,1217.000000,594.503284],
[4.500000,1183.886353,2099.999905],
[9.000000,1220.000000,1071.599126],
[13.500000,1184.565430,2099.999905],
[18.000000,1219.000000,435.631812],
[22.500000,1185.150635,2099.999905],
[24.500000,1217.555542,320.541441],
[27.000000,1185.427490,2099.999905],
[31.500000,1216.000000,300.000012],
[36.000000,1185.981445,2099.999905],
[40.500000,1215.000000,306.778669],
[45.000000,1186.629272,2099.999905],
[49.500000,1216.000000,300.000012],
[54.000000,1187.214478,2099.999905],
[58.500000,1215.000000,300.000012],
[63.000000,1187.893555,2099.999905],
[67.500000,1218.000000,335.902870],
[72.000000,1188.572510,2099.999905],
[76.500000,1220.000000,359.615386],
[81.000000,1189.282715,2099.999905],
[85.500000,1224.000000,1382.480264],
[90.000000,1189.992920,2099.999905],
[94.500000,1225.000000,1206.023455],
[99.000000,1190.578125,2099.999905]])
#upper,limit
h = np.array([[0.0,1217],
[4.5,1217],
[9.0,1220],
[13.5,1219],
[18.0,1219],
[22.5,1218],
[27.0,1217],
[31.5,1216],
[36.0,1215],
[40.5,1215],
[45.0,1216],
[49.5,1216],
[54.0,1215],
[58.5,1215],
[63.0,1217],
[67.5,1218],
[72.0,1219],
[76.5,1220],
[81.0,1222],
[85.5,1224],
[90.0,1225],
[94.5,1225],
[99.0,1224],
[103.5,1225]])
##
x=M[:,0]
y=M[:,1]
z=M[:,2]
##
xi=np.linspace(0,100.,100.)
yi=np.linspace(1190,1225.,1225.)
X,Y= np.meshgrid(xi,yi)
Z = griddata((x, y), z, (X, Y),method='cubic')
plt.contourf(X,Y,Z)
plt.colorbar()
plt.plot(h[:,0],h[:,1],'black',linewidth=2)
plt.scatter(x,y,marker ='o',c='k',s=10,zorder=10)
plt.xlim(0,100)
plt.ylim(1190,1225)
plt.grid(True)
plt.rc('grid',linestyle="-",color='black')
plt.show()
想法是隐藏或删除黑线上方的轮廓图。我用一个补丁在Matlab上解决了这个问题，想知道用python解决它的最佳方法吗？
答案 0 :(得分：0)
我不确定您是否可以使用现有功能来实现此目的。
您可以做的是对网格的所有点进行检查，如果y超过h所定义的极限，则将X，Y，Z值命名为：
（分配Z后添加这段代码）
x0 = 0
dx = 4.5
for n1, xv in enumerate( X):
for n2, x in enumerate( xv):
index = (x - x0) / dx
i1, i0 = int(np.ceil( index)), int(np.floor( index))
y = (h[i1][1] - h[i0][1]) / dx * (x - h[i0][0]) + h[i0][1]
if Y[n1][n2] > y:
X[n1][n2] = np.nan
Y[n1][n2] = np.nan
Z[n1][n2] = np.nan
编辑：添加了完整代码，以避免复制粘贴问题
from scipy.interpolate import griddata
import numpy as np
import matplotlib.pyplot as plt
M = np.array([[0.000000,1217.000000,594.503284],
[4.500000,1183.886353,2099.999905],
[9.000000,1220.000000,1071.599126],
[13.500000,1184.565430,2099.999905],
[18.000000,1219.000000,435.631812],
[22.500000,1185.150635,2099.999905],
[24.500000,1217.555542,320.541441],
[27.000000,1185.427490,2099.999905],
[31.500000,1216.000000,300.000012],
[36.000000,1185.981445,2099.999905],
[40.500000,1215.000000,306.778669],
[45.000000,1186.629272,2099.999905],
[49.500000,1216.000000,300.000012],
[54.000000,1187.214478,2099.999905],
[58.500000,1215.000000,300.000012],
[63.000000,1187.893555,2099.999905],
[67.500000,1218.000000,335.902870],
[72.000000,1188.572510,2099.999905],
[76.500000,1220.000000,359.615386],
[81.000000,1189.282715,2099.999905],
[85.500000,1224.000000,1382.480264],
[90.000000,1189.992920,2099.999905],
[94.500000,1225.000000,1206.023455],
[99.000000,1190.578125,2099.999905]])
#upper,limit
h = np.array([[0.0,1217],
[4.5,1217],
[9.0,1220],
[13.5,1219],
[18.0,1219],
[22.5,1218],
[27.0,1217],
[31.5,1216],
[36.0,1215],
[40.5,1215],
[45.0,1216],
[49.5,1216],
[54.0,1215],
[58.5,1215],
[63.0,1217],
[67.5,1218],
[72.0,1219],
[76.5,1220],
[81.0,1222],
[85.5,1224],
[90.0,1225],
[94.5,1225],
[99.0,1224],
[103.5,1225]])
##
x=M[:,0]
y=M[:,1]
z=M[:,2]
##
xi=np.linspace(0,100.,100.)
yi=np.linspace(1190,1225.,1225.)
X,Y= np.meshgrid(xi,yi)
Z = griddata((x, y), z, (X, Y),method='cubic')
x0 = 0
dx = 4.5
for n1, xv in enumerate( X):
for n2, x in enumerate( xv):
index = (x - x0) / dx
i1, i0 = int(np.ceil( index)), int(np.floor( index))
y = (h[i1][1] - h[i0][1]) / dx * (x - h[i0][0]) + h[i0][1]
if Y[n1][n2] > y:
X[n1][n2] = np.nan
Y[n1][n2] = np.nan
Z[n1][n2] = np.nan
plt.contourf(X,Y,Z)
plt.colorbar()
plt.plot(h[:,0],h[:,1],'black',linewidth=2)
plt.scatter(x,y,marker ='o',c='k',s=10,zorder=10)
plt.xlim(0,100)
plt.ylim(1190,1225)
plt.grid(True)
plt.rc('grid',linestyle="-",color='black')
plt.show()